 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1999 The Journal of Bone and Joint Surgery, Inc. Impaction Allografting with Cement for Revision of the Femoral Component. A Minimum Four-Year Follow-up Study with Use of a Precoated Femoral Stem* , CHICAGO, ILLINOISInvestigation performed at the Department of Orthopaedic Surgery, Rush-Presbyterian-St. Luke's Medical Center, Chicago
Background: Cancellous impaction allografting with cement for revision of the femoral component has conventionally been performed with a polished, tapered implant, which was designed to allow subsidence of the component. However, subsidence has been associated with pain in the thigh, dislocation of the hip, and revision of the component. This prospective study tested the hypothesis that good clinical results can be achieved—without subsidence of the component—with use of impaction allografting and a precoated, collared, straight stem for difficult femoral revisions. Methods: Twenty-nine patients had revision of the femoral component with use of impaction allografting with cement and a Harris Precoat stem. Impaction allografting was performed when loss of metaphyseal and diaphyseal bone precluded revision with more straightforward techniques or when reconstitution of bone was considered a specific goal of the reconstruction (as was sometimes the case with revision of the component in younger patients). The patients were followed prospectively and were evaluated with use of the Harris hip score and serial radiographs. The patients were followed for a minimum of four years (mean, sixty-three months), except for four who died. Results: Four patients died before the minimum four-year follow-up period had elapsed; all four had the prosthesis in place at the time of death. The Harris hip scores improved from a preoperative mean of 54 points (poor) (range, 21 to 91 points) to a mean of 87 points (good) (range, 41 to 100 points) at the time of the most recent follow-up. Kaplan-Meier survivorship analysis, with aseptic loosening as the end point, was 92 percent (95 percent confidence interval, 82 to 100 percent) at six years; one additional hip failed because of a hematogenous infection at seventy-three months postoperatively, for an overall failure rate of 12 percent (three of twenty-five patients) at the time of the most recent follow-up. Two hips needed a repeat revision; one was revised because of subsidence of the stem with recurrent osteolysis and the other, in a patient who had hemodialysis, because of late sepsis. A third femoral component subsided and failed but was not revised. Radiographic evidence of bone-stock reconstitution was observed in six (29 percent) of the twenty-one patients for whom radiographs were available. As in other series of patients managed with impaction allografting, the complication rate was high; excluding the revisions, three reoperations were performed, and six patients had either intraoperative femoral fracture or perforation necessitating cerclage wiring or cortical strut allografting and cerclage wiring at the time of the procedure. There were six nonunions in eighteen patients who had been operated on with a transtrochanteric approach. Conclusions: Difficult revisions of the femoral component with use of impaction allografting and a precoated stem provided satisfactory clinical and radiographic results at the time of intermediate-term follow-up. However, the high rate of complications in our series led us to refine our indications for the procedure.
Loosening of the femoral component after total hip arthroplasty is often associated with loss of proximal femoral bone stock. In this setting, some revision techniques are performed with the goal of providing a stable and durable reconstruction despite the loss of bone13,22,24,35,37,40,41, whereas others—including impaction allografting—are done in an attempt to reconstitute the bone-deficient proximal aspect of the femur as well1,8,10,11,16,32,42. Reports from several centers have presented the short-term follow-up results of cancellous impaction grafting with cement for revision of the femoral component. Early results were encouraging11,16, but recent studies from other centers demonstrated a high rate of massive, symptomatic subsidence10 and frequent perioperative femoral fractures32, both during and after the operation. In all of the clinical series of which we are aware, this technique has been performed with the Exeter stem (Howmedica, Rutherford, New Jersey, and Howmedica International, Staines, Middlesex, England) or the CPT (collarless, polished, tapered) stem (Zimmer, Warsaw, Indiana), both of which are double-tapered, highly polished, collarless devices10,11,16,30,32. Advocates for the use of these devices with this technique have stated that subsidence in this setting does not lead to clinical failure16. In fact, subsidence of these wedge-shaped stems is said to be necessary to load the bone graft properly11,16. However, a previous study of femoral components inserted with more traditional means of cement fixation suggested that subsidence is detrimental20, and the results of two other clinical series raised concerns about the frequency and amount of subsidence of double-tapered stems inserted after impaction grafting10,32. The goals of impaction grafting include reconstitution of bone stock and stable fixation. If these goals can be met with a device that was not designed to subside, use of such a stem seems potentially advantageous. This prospective study tests the hypothesis that good clinical results can be achieved with difficult revisions of the femoral component involving use of cancellous impaction allografting and a precoated, collared, straight stem that has been used extensively for femoral reconstruction with cement4,33,39. The results were evaluated after a minimum of four years of follow-up.
Patients Four patients died before the minimum four-year follow-up period had elapsed: one died of pulmonary embolism at six weeks postoperatively and three, of unrelated causes at a mean of twenty-six months postoperatively. Of the twenty-five patients who were followed for at least four years, twenty-one both had radiographs made and had a clinical examination that included use of a standardized hip form and the remaining four were interviewed by telephone. These four patients did not return to our clinic for radiographic and clinical examination despite substantial efforts to have them do so. The mean duration of follow-up of the patients who were alive at four years was sixty-three months (range, forty-eight to eighty-two months). In the entire cohort of twenty-nine patients, the mean age at the time of the operation was sixty-four years (range, thirty-three to eighty-three years) and the mean number of previous operations on the affected hip was 1.6 (one, two, or three). The preoperative diagnosis was aseptic loosening of the femoral component in twenty-seven patients. The other two patients had the index procedure as a second-stage reimplantation after previous removal of the femoral component because of infection. Patients who needed revision of the femoral component were included in our study if it was desirable to reconstitute proximal bone stock (as in a younger patient who had severe loss of metaphyseal bone) or if press-fitting into the remaining endosteum without cement would have required an overly large stem. All patients had the procedure performed with the same operative technique and the same type of femoral component by one of the four of us who are senior staff surgeons (R. A. B., A. G. R., J. J. J., and J. O. G.). During the twenty-nine-month period of this study, the four of us who are surgeons performed a total of eighty-seven revisions of the femoral component. A standard cementing technique was used in thirty-one of those revisions; a fully porous-coated press-fit stem (AML; DePuy, Warsaw, Indiana), in nineteen (two of which involved use of a component that was more than eighteen millimeters in diameter); an allograft-prosthetic composite, in three; a proximally coated press-fit femoral stem, in three; and impaction grafting with a Harris CDH stem (Zimmer), in two. Thus, 33 percent (twenty-nine) of the eighty-seven revisions of the femoral stem performed during that time involved impaction grafting and a standard precoated stem. The technique was used for patients who had more severe femoral bone-stock deficiency.
Prostheses and Operative Technique Body-exhaust suits and operating rooms with vertical laminar airflow were used for all procedures. Antibiotic prophylaxis was used for all patients once specimens had been obtained for intraoperative culture. A standard transtrochanteric approach was used in eighteen (62 percent) of the twenty-nine patients, and a posterior approach was used in the other eleven (38 percent).
Cancellous Impaction Grafting
Postoperative Care
Clinical and Radiographic Analysis That observer used the Endo-Klinik system12, a simple four-grade descriptive classification employed in what we believe to be the first published study of impaction grafting16, to grade the femoral bone stock on the basis of the preoperative radiographs. According to the Endo-Klinik scale, two (7 percent) of the patients had grade-I bone stock (aseptic loosening without widening of the intramedullary cavity), eighteen (62 percent) had grade-II (endosteal erosion and mild widening of the intramedullary cavity), eight (28 percent) had grade-III (expansile widening of the proximal part of the femur), and one (3 percent) had grade-IV (gross destruction of the proximal part of the femur with severe diaphyseal widening). The same observer also analyzed all postoperative radiographs. The cementing technique was graded according to accepted criteria3,37. The most recent radiographs were evaluated for signs of loosening of the component, and subsidence and changes in the position of the component were measured with a digitizing tablet (Sigma Scan; Jandel Scientific, Corte Madera, California) as previously described29. Radiographs were evaluated for evidence of lucency or debonding at the stem-cement interface, fracture of the cement, osteolysis, and a complete radiolucent line at the bone-cement or graft-host interface on both anteroposterior and lateral radiographs. New or progressive heterotopic ossification was graded according to the classification system of Brooker et al.7. The criteria of Gie et al.16 were used to grade incorporation of the cancellous allograft. Because of concerns about interobserver and intraobserver reliability23, the radiographs were also evaluated with a simpler method. The appearance of the cancellous graft on the most recent follow-up radiographs was compared with the postoperative appearance and described as either unchanged, less radiopaque, or more radiopaque; the cortical bone stock was characterized as unchanged, hypertrophied, or atrophied. Care was taken to account for differences between radiographic technique, position, and degree of penetration so that these would not be misinterpreted as changes in the status of the bone graft or cortex. The appearance of the cortical bone distal to the implant and the cancellous bone in the pelvis a distance from the implant were used to control for these differences in technique.
Statistical Methods
Clinical and Radiographic Results Of the twenty-five patients who were still living at a minimum of four years postoperatively, twenty-two (88 percent) had a well functioning prosthesis. Two patients had a repeat revision. One revision was performed at fifty-three months because of recurrent femoral osteolysis and painful subsidence of the stem (Figs. 1-A, 1-B, 1-C, 1-D), and the other (a resection arthroplasty) was done at seventy-three months because of hematogenous sepsis. A third patient had painful subsidence and clinical failure at twenty-six months. In that patient, the femur had had a varus deformity at the junction of the metaphysis and the proximal part of the diaphysis, but an osteotomy had not been done before the index revision of the femoral component (Figs. 2-A, 2-B, and 2-C). The patient did not have a subsequent revision because of poor health. Kaplan-Meier analysis21, with aseptic loosening as the end point, revealed a rate of survival of 92 percent (95 percent confidence interval, 82 to 100 percent) at six years (Fig. 3-A). The overall failure rate for the entire series, taking into account both revision for any reason and radiographic evidence of subsidence, was 12 percent (three of twenty-five) (Fig. 3-B). Neither of the two patients in whom this procedure was performed as a second-stage reimplantation after removal of the component because of infection had had a recurrent infection at the time of the most recent follow-up.
The Harris hip scores improved from a mean of 54 points (range, 21 to 91 points) preoperatively to a mean of 87 points (range, 41 to 100 points) at the time of the most recent follow-up (p < 0.0005). Preoperatively, seven (24 percent) of the twenty-nine patients rated pain as none or slight; at the most recent follow-up evaluation, twenty-two (96 percent) of the twenty-three patients who had not had a revision and were alive rated pain as none or slight. Similarly, twenty-three (79 percent) of the twenty-nine patients used a cane or crutches preoperatively, whereas nine (39 percent) of the twenty-three did so at the time of the most recent follow-up. Twenty-one (91 percent) of the twenty-three patients who had not had a revision and had been followed for at least four years had a good or excellent result, defined as a Harris hip score of at least 80 points (Table I). The patient in whom the stem failed but was not revised was included in all analyses.
Of the twenty-nine patients, none had grade-A cementing technique3 (so-called whiteout), one (3 percent) had grade-B, one had grade-C1, fourteen (48 percent) had grade-C2, and thirteen (45 percent) had grade-D. Review of the radiographs available at the time of the most recent follow-up revealed that nineteen (90 percent) of the twenty-one femoral components were stable, without any measurable subsidence. Apart from the two stems with known aseptic loosening, no stem had visible radiolucency or debonding at the stem-cement or cement-bone interface, fracture of the cement, or visible femoral or pelvic osteolysis. With care taken to account for differences in radiographic technique, it was determined that six (29 percent) of the twenty-one hips demonstrated some reconstitution of femoral bone stock or visible incorporation of the graft, twelve (57 percent) showed neither reconstitution nor progressive loss of bone stock, and the remaining three (14 percent) had continued loss of cortical or cancellous bone (Table II). With the numbers available, no association could be detected between the preoperative Endo-Klinik grade12 or the grade of the cementing technique3 and any radiographic or clinical end point (reconstitution of bone stock, subsidence of the stem, or the hip score). No patient had progressive heterotopic ossification, defined as an increase of one grade7 or more compared with the radiographic appearance before the index revision.
Complications There were four early postoperative complications. These included two nerve palsies (one sciatic and one peroneal), neither of which resolved completely; one hip dislocation at six months, which was treated closed and had not recurred by the time of the most recent follow-up; and one compartment syndrome, in the patient who had the sciatic nerve palsy. The compartment syndrome involved the contralateral gluteal compartment following the six-hour operation in the lateral position. It was diagnosed in the immediate postoperative period on clinical examination and with direct manometry, and an emergency fasciotomy was performed. The patient subsequently had myoglobinuria and acute renal failure requiring a short period of dialysis but eventually had full recovery of renal function. There were two other reoperations, in addition to the revision and the resection arthroplasty. One patient was managed with removal of allograft struts and a quadricepsplasty at two months postoperatively because of stiffness of the knee, and another had removal of a trochanteric cable-grip at eighteen months because of bursitis. Six of the eighteen patients managed with trochanteric osteotomy had nonunion with migration. One of these patients had the single episode of hip dislocation described earlier. Another was the patient who had the reoperation for the removal of painful trochanteric hardware. At that reoperation, the fibrous trochanteric nonunion was found to be stable. In five of the six patients who had a trochanteric nonunion, debris from the braided cables used for trochanteric fixation was visible in the joint radiographically. One of those patients had severe polyethylene wear, osteolysis, and revision of the femoral component because of subsidence.
Appearance of the Allograft at the Repeat Revision
We performed a prospective clinical study to evaluate the results of revision of a femoral component with use of impaction allografting and a textured, precoated, collared stem. The hypothesis that this technique can produce satisfactory clinical results with a stem that is neither polished and tapered nor designed to permit subsidence seemed to be valid after a minimum of four years (mean, sixty-three months) of follow-up. Four years of follow-up is too short to determine whether this reconstructive technique will ultimately succeed. At this time, we can only compare its early results with those of established techniques to verify that it does not result in early failure. In the present series, which represents our initial experience with this technically demanding procedure, the rate of prosthetic survival, according to Kaplan-Meier analysis21, was 92 percent (95 percent confidence interval, 82 to 100 percent) at six years when the end point was aseptic loosening, and the overall survival rate was 88 percent (twenty-two of twenty-five hips) when the one hematogenous infection was included in the analysis. In addition, twenty-one (91 percent) of the twenty-three patients who had not had a revision and had been followed for at least four years had a good or excellent clinical result (a Harris hip score of at least 80 points). These findings are comparable with the results after similar durations of follow-up in the best contemporary series of revisions performed with cement, in which rates of loosening and failure were reported to be 7 percent (twenty-five of 351)40 and 10 percent (five of forty-nine)41 at six years. In addition, it did not appear that patients who had bone-stock deficiency had been preselected in those studies as they had been in ours (Figs. 4-A, 4-B, and 4-C). Whether revision of the femoral component with impaction grafting will prove durable over the long term remains to be seen.
Previous studies of impaction allografting of which we are aware have demonstrated rates of revision ranging from 3 percent [one of thirty-five]30 to 10 percent (eight of seventy-nine)10. However, the mean durations of follow-up in those reports were quite short, with none longer than thirty-one months10,11,16,30,32. In addition, variation in the criteria for inclusion of patients in studies of impaction grafting renders interpretation of the literature more complicated, and direct comparisons of failure rates are probably misleading. One of the coauthors of the original report on impaction grafting expressed reservations about using the technique for the most challenging cases26, which was what was done in our series and others30,32. Considering the longer follow-up and stricter inclusion criteria used in the present study, the rate of revision and the clinical results compare favorably with those of other reports (Table III). For example, the proportion of patients in this study who had severe bone-stock deficiency (grades III and IV according to the Endo-Klinik scale12) is more than double that in the study by Gie et al.16. In another series11, more than one-third (nineteen) of the fifty-six patients had a minimally damaged (grade-I12) femur compared with only 7 percent (two) of the twenty-nine patients in the present series.
As far as we know, all other studies of impaction allografting10,11,16,30,32 have involved stems that differed substantially, with regard to surface and geometry, from the stems used in our series. Both the Exeter and the CPT stems are collarless, polished, and tapered, and they are thought to allow subsidence within the cement mantle without causing symptomatic loosening16. Several millimeters of subsidence is tolerated by some designs of stems when they are used for primary total hip arthroplasty performed with cement27,28, but more severe subsidence typically results in failure5. The precoated stem that was used in the present series was designed to achieve initial stability6 and a bond with the cement-bone graft composite4,33,39. Unlike the situation with polished stems, any debonding or subsidence of precoated stems leads to failure of a primary total hip arthroplasty performed with cement34. In the present series, this was found to be true for revision arthroplasties performed with impaction allografting; both components that subsided caused symptoms. The contention that subsidence is desirable when impaction grafting is performed with a polished, tapered stem11,16 has been called into question. It was believed that these devices subsided because of so-called cold flow of the cement, allowing the wedge-shaped stem to become more tightly fixed as it subsided11,14,16. More recent reports have shown that this does not always occur. Masterson et al.30 found a high prevalence of early fracture of the cement around stems that subsided, an observation that does not support the hypothesis that cold flow is the mechanism of subsidence. In another study, roentgen stereophotogrammetric analysis was used to arrive at the same conclusion15. Subsidence may be massive10,30, and it may continue for an unpredictable period of time postoperatively32; such subsidence has been convincingly associated with pain in the thigh10 and prosthetic dislocation30. Achieving a circumferential cement mantle is important to the success of a femoral stem inserted with cement9,36, and an inadequate cementing technique appears to be one cause of subsidence of stems inserted after impaction allografting30,31; however, obtaining a consistent cement mantle with impaction grafting is difficult. Masterson et al.30,31 found that, when stems had been inserted after use of the original Exeter rasps, which were not oversized, 40 percent of the zones described by Gruen et al.18 had no visible cement. Cementing of the Precoat and CPT systems, both of which are implanted after the use of rasps that are larger than the final implant, left complete voids in 14 and 15 percent of the zones, respectively31. In the present study, we did not perform an analysis according to the zones of Gruen et al. but instead used a clinically validated system of grading the cementing technique on the basis of good-quality anteroposterior and lateral radiographs for each patient3,37. This approach is somewhat stricter than a zonal analysis because even one large void results in a grade-D mantle. As assessed with this method, 93 percent (twenty-seven) of the twenty-nine hips had an unsatisfactory cement mantle (grade C2 or D). This finding highlights the technical difficulty of obtaining a consistent cement mantle with impaction allografting, even when specially designed, oversized rasps are used. However, the presence of morseled allograft pieces in the cement may render analysis of these hips with conventional methods—such as the zonal analysis30,31 or the technique used in our study3,37—inadequate or misleading. In most of the clinical studies of impaction grafting of which we are aware, histological healing or incorporation of the cancellous allograft was inferred from plain radiographs11,16,32, despite the poorly established relationship between apparent radiographic remodeling of cancellous allograft and the actual viability of the graft25,38. In the present study, we found convincing evidence of bone-graft remodeling in only six (29 percent) of twenty-one patients at the time of the most recent follow-up; this rate is much lower than that reported in other series16,32. It is not possible to determine from these data whether the low rate of radiographic evidence of bone-remodeling in our series represents an actual difference in the biology of the stem-graft-host composite or simply a stricter interpretation of the radiographs. The intraoperative appearance of the graft-host junction in the two patients who had a repeat revision, as well as the fact that all but two stems remained radiographically stable after a mean duration of follow-up of more than five years, supports the contention that bone-healing is taking place but is not seen on plain radiographs, as has been reported elsewhere38. The prevalence of serious complications and reoperations in the present series of twenty-nine patients was high: six patients (21 percent) had an intraoperative femoral fracture or perforation, and three (10 percent) had a reoperation other than a revision. The technique is also time-consuming, technically demanding, and expensive. On the basis of these issues and reports that have documented excellent intermediate and long-term results with large, extensively coated femoral stems inserted without cement2,13,24, we have narrowed our indications for impaction allografting. In our opinion, impaction allografting is most appropriate for a patient who needs revision of the femoral component and has severe loss of bone from the proximal aspect of the femur, a patulous or severely ectatic proximal part of the diaphysis, thin cortices, and a large-diameter canal (typically more than eighteen millimeters). In such patients, distal fixation without cement may be compromised by an inability to obtain rotational stability of the prosthesis and the potential for proximal stress-shielding, and traditional revision with cement might be unsatisfactory because of the condition of the remaining endosteum. There are a number of variables that may affect the outcome of a revision of the femoral component with impaction allografting. Series in which similar implants and similar inclusion criteria were used may still differ with respect to the cement (the technique, the type, or the viscosity), the allograft (the source, the consistency, or whether there was pretreatment with radiation or freeze-drying), the operative approach, and the postoperative care, to name but a few potentially important factors. The effects of these variables on the clinical results of impaction grafting are, for the most part, not yet known. NOTE: The authors thank Susan Shott, Ph.D., for statistical analysis; Paul Sauer, M.D., for preliminary radiographic analysis; and Karen Kiafeta, B.S., for technical assistance.
*Although none of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article, benefits have been or will be received, but are directed solely to a research fund, foundation, educational institution, or other nonprofit organization with which one or more of the authors is associated.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |