 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1996 The Journal of Bone and Joint Surgery, Inc. A Tapered Titanium Femoral Stem Inserted without Cement in a Total Hip Arthroplasty. Radiographic Evaluation and Stability*   , LONDON, ONTARIO, CANADAInvestigation performed at the Division of Orthopaedic Surgery, University Hospital, London
The results of 416 total hip arthroplasties performed in 372 patients with insertion of a Mallory-Head titanium femoral component without cement were reviewed retrospectively, to determine the stability and the radiographic parameters of stability for this type of stem. The average duration of clinical and radiographic follow-up was 3.7 years (range, 2.0 to 6.5 years). No revisions were performed to treat instability of the stem, pain, or osteolysis by the time of the latest follow-up examination. Although the clinical results were excellent, the radiographic signs of fixation that have been well described for chromium-cobalt stems were frequently absent. So-called endosteal spot welds, commonly seen in association with long, extensively coated stems, were seen infrequently. Resorptive bone-remodeling rarely extended beyond zones 1 and 7 of Gruen et al. Both endosteal and periosteal distal cortical hypertrophy was common, but the cause was unknown. Osteolysis of the distal aspect of the femur, shedding of the porous coating, and breakage of the stem were not seen. Forty-four stems (11 per cent) had initial subsidence; however, it nearly always stabilized within six months postoperatively and did not affect the eventual stability at the time of the short-term follow-up. The short-term clinical outcome confirmed the stability of these stems that had been determined from the radiographic findings and the lack of revisions. At the latest follow-up examination, 323 (87 per cent) of the patients (362 [87 per cent] of the hips) had no or slight pain in the lower limb over-all and only twenty-two (6 per cent) of the patients (twenty-seven [6 per cent] of the hips) had pain in the thigh. This tapered titanium femoral component appears to provide excellent short-term stability, without resorptive bone-remodeling, osteolysis, or a high prevalence of pain in the thigh. Radiographic evaluation of stability of this stem is clearly different from that of chromium-cobalt stems—particularly extensively coated anatomic medullary locking stems—in that so-called spot welds and severe resorptive bone changes were uncommon and distal cortical hypertrophy was common.
Early reports of total hip arthroplasties performed with cement predicted high rates of long-term failure69,73, which led to efforts to improve fixation with cement and to achieve fixation without cement. Improved techniques for the application of cement, including bone-cleaning and pressurization of the cement, have clearly improved the radiographic results and the longevity of cemented stems30,31,54,62. Fixation without cement has most commonly been attempted with use of porous surfaces for bone ingrowth. It was hoped that these surfaces would provide durable biological fixation with avoidance of late loosening and osteolysis (so-called cement disease)40. However, the results of early total hip arthroplasties performed without cement often have not been as good as those of arthroplasties performed with newer cementing techniques42,49,51. In addition, stems inserted without cement have been associated with early and late instability11,42-44,51,68, persistent pain in the thigh5,6,8,11,15,20,27,34,36,41,60,66, and osteolysis43,50,52,63,74,77. One femoral component designed for insertion without cement, the anatomic medullary locking stem (DePuy, Warsaw, Indiana), has been associated with excellent intermediate-term clinical and radiographic results19,21. On the basis of well described and accepted radiographic criteria, stability can be achieved in most patients22. However, stress-shielding leading to resorptive bone-remodeling has been a major concern, with the potential for failure of the implant or the femur and difficulty with the performance of a revision operation18,23,24,35,58,70. We have had a good deal of experience with the Mallory-Head total hip prosthesis (Biomet, Warsaw, Indiana), which includes a tapered titanium stem designed to be inserted without cement. Excellent clinical results have been achieved despite the lack of typical radiographic signs of stable fixation that have been described for the anatomic medullary locking stem and other chromium-cobalt stems22. The present study describes the stability and the radiographic signs of stability of the Mallory-Head femoral stem in a large series of patients who had been followed for at least two years.
The straight Mallory-Head stem is constructed of a titanium alloy and has a 3-degree taper in both the coronal and the sagittal plane. The proximal third has a plasma-spray porous coating with a gradient of porosity and pore size—that is, the pore size and the over-all degree of porosity increase from the substrate peripherally2. The middle third has a textured surface, and the distal third is smooth. There are proximal-lateral, anterior, and posterior flanges to enhance fixation in the metaphysis, as well as a small collar. The stems are available in diameters ranging from six to seventeen millimeters and in progressively increasing lengths ranging from 135 to 170 millimeters. A variety of neck lengths are also available. A twenty-eight-millimeter femoral head was used in all but ten of the 372 patients, who received a twenty-two-millimeter head for congenital dislocation of the hip. Total hip arthroplasty with the Mallory-Head prosthesis and without cement has been done at our institution since October 1987. A total of 442 such primary total hip arthroplasties were performed between October 1987 and January 1992. Eighteen patients (nineteen hips) died, and five patients (five hips) were lost to follow-up before the two-year review. None of these hips were included in the final analysis. Two femoral components that were revised, one for deep infection and the other for non-union at the site of a concomitant femoral osteotomy in a patient who had congenital dislocation of the hip, were also excluded from the analysis as the complications were not considered directly attributable to the stability of the stem. The remaining 416 hips (372 patients) were followed clinically and radiographically for a minimum of two years. Two years was selected as the minimum duration of follow-up because, with other types of stems, bone-remodeling has rarely occurred after this period18, and our experience with this stem has been the same. The average duration of follow-up was 3.7 years (range, 2.0 to 6.5 years). Three hundred and ten hips were followed for at least three years; 227, for at least four years; 107, for at least five years; and twenty-seven, for at least six years. Two hundred and eighteen prostheses were implanted in female patients and 198, in male patients. Two hundred and sixteen were inserted on the left side and 200, on the right. The average age of the patients was sixty-three years (range, seventeen to eighty-three years) at the time of insertion of the prosthesis and sixty-seven years (range, nineteen to eighty-five years) at the time of the latest follow-up. The arthroplasty was performed to treat osteoarthrosis (including that with a traumatic or childhood etiology) in 341 hips (82 per cent), rheumatoid arthritis in twenty-eight (7 per cent), osteonecrosis in twenty-four (6 per cent), congenital dislocation in thirteen (3 per cent), and another diagnosis in ten (2 per cent). We are reporting on a selected series in that approximately 240 total hip arthroplasties with cement and hybrid total hip arthroplasties (one component inserted with cement and the other inserted without it) were performed at our institution during the same period. One hundred and twenty-five of the total hip arthroplasties with cement were performed as part of a randomized clinical trial. The remainder of the arthroplasties with cement and hybrid arthroplasties were performed in patients who were of advanced age or who requested cementing of the femoral component. However, the patients who had a Mallory-Head implant inserted without cement varied widely with regard to age, diagnosis, and quality of bone, and they are included in the present analysis. The technique for insertion involves use of transparent templates on preoperative radiographs to determine the approximate diameter of the stem and the appropriate level of resection of the femoral neck. The use of manual reamers is followed by the use of broaches, but no milling or machining of the femur is needed. The diameter of the implant inserted is the same as that of the largest reamer and the size is the same as that of the largest broach used. Intraoperative radiographs are not routinely made at our institution. All operations were performed through a modified direct lateral approach at one institution, under the direct supervision of two of us (R. B. B. and C. H. R.), senior surgeons. Laminar airflow, prophylactic antibiotics, and body-exhaust suits were routinely used. Coumadin (warfarin) was generally used as postoperative prophylaxis against thromboembolism. Physiotherapists directed the postoperative rehabilitation, which included restriction to 50 per cent weight-bearing on crutches for six weeks, with progression to full weight-bearing and walking with a cane for another six weeks. Hip abduction exercises were performed throughout this period, with use of gradually increasing resistance. Supervision by the physiotherapist was generally discontinued at three months postoperatively. Patients were seen postoperatively at six weeks; at three, six, and twelve months; and yearly thereafter. Two hip-rating systems were used during this study. Information was initially recorded with the Harris hip-rating system29, but The American Academy of Orthopaedic Surgeons/Hip Society questionnaire was instituted following the recommendation of Johnston et al.39. Because of this transition and the known lack of uniformity of hip-rating scores9, only the individual parameters of pain and limp are reported. Patients were asked two questions regarding pain. The first question referred to over-all pain in the hip or the lower limb and is reflected in the pain score (with a total possible score of 44 points). Second, the patients were queried specifically about pain in the thigh, which they rated on a visual-analog scale of 1 to 10. For reporting purposes, a score of 1 to 4 indicated mild pain; 5, 6, or 7, moderate pain; and 8, 9, or 10, severe pain. There was no attempt to subdivide the patients into Charnley functional classes12. Clinical information, including complications, was recorded prospectively in a computer database. Anteroposterior radiographs of the pelvis and cross-table lateral radiographs of the hip were made preoperatively, immediately postoperatively, and at each follow-up visit. Radiographic analysis was performed by two of us (B. D. M. and N. N.), orthopaedic fellows, without knowledge of the clinical results. Two measures of the preoperative quality of the bone were used. The cortical index was measured as the ratio of the cortical diameter to the medullary diameter ten centimeters distal to the lesser trochanter (Fig. 1). This is a good measure of the quality of cortical bone and is the critical factor in the prediction of postoperative bone resorption when a stem fills the femoral canal18. There is a strong association between the cortical index and the calcar-to-canal isthmus ratio of Dorr16, which can be difficult to measure13,17. The modified trabecular index of Singh et al. was also recorded67. Both indices were arbitrarily graded in an attempt to classify the quality of bone.
Postoperatively, the alignment of the femoral stem was referenced with respect to the axial alignment of the proximal aspect of the femur on the anteroposterior radiograph. Alignment was considered neutral if the angle between the two reference lines was 3 degrees or less from collinearity (Fig. 2). Contact of the collar with the calcar of the femur was recorded as present or absent. The percentage of the metaphysis occupied by the stem (the metaphyseal fit) was measured at the lesser trochanter and was considered good if the stem occupied 80 per cent of the canal on the anteroposterior radiograph and 70 per cent on the lateral radiograph, excluding the trochanters. A similar measurement was done to determine the isthmal fill (the degree to which the implant filled the medullary canal of the diaphysis) three centimeters proximal to the tip of the implant. Three-point contact between the stem and the endosteal cortex, on either the anteroposterior or the lateral radiograph, was also considered to indicate good isthmal fill.
Serial radiographs, including those made at the latest follow-up examination, were available for all patients. Several definitions and measurements were used. Calcar rounding was defined as rounding of the calcar of the femur within six months postoperatively. Resorptive bone-remodeling, if present, was classified as mottling if there was loss of trabecular bone density or corticocancellization (a blending of cortical and cancellous bone into a uniform density) or it was classified as cortical thinning. So-called endosteal spot welds were defined as new bone bridging the endosteum and the porous surface of the implant. Cortical hypertrophy was defined as new bone of cortical density resulting in increased cortical thickness and was classified as either endosteal or periosteal. These bone-remodeling changes, and any radiolucent or radiodense lines, were assessed with use of the zones of Gruen et al.26 on the anteroposterior radiographs and with use of the zones of Johnston et al.39 on the lateral radiographs. Generally, the radiographic findings were more subtle on the lateral radiographs and only confirmed those seen on the anteroposterior radiographs. Therefore, only the findings in the anteroposterior zones are reported, for simplicity. (For example, cortical hypertrophy seen in anteroposterior zones 3 and 5 and lateral zones 10 and 12 is reported as occurring in zones 3 and 5.) A halo pedestal was defined as thin radiodense lines surrounding the tip of the implant. An incomplete shelf of dense bone at the tip of the stem was called a shelf pedestal. Subsidence was measured with use of a perpendicular line drawn from the greater trochanter to the lateral border of the implant as well as from the collar to the lesser trochanter as references (Fig. 2). A difference of three millimeters in both measurements was considered important and measurable. Any shedding of the porous coating, breakage of the stem, or osteolysis (defined as a new area of cystic radiolucency or endosteal erosion) was recorded. We had intended to define instability of the Mallory-Head stem with use of previously accepted criteria for evaluation of chromium-cobalt stems. However, as it became clear that many of these radiographic features (for example, spot welds) were not typically present, the criteria were modified. A stem was therefore considered unstable if there was progressive subsidence or a circumferential radiolucent or radiodense line. In the evaluation of other types of stems, subsidence is considered a major sign of instability of the implant and the absence of reactive lines around the porous surface is considered a major sign of failed osseous integration. These two findings have been persistently associated with unstable femoral components, similar to the Mallory-Head stem, used in revision total hip arthroplasty at our institution53. We attempted to determine if preoperative and postoperative radiographic variables were associated with the eventual clinical and radiographic outcomes. The chi-square test was used when both variables were categorical, and analysis of variance was used when one variable was continuous.
Clinical Results As rated with use of the Harris hip score, the average pain score at one and two years postoperatively was 41 of 44 points; 90 per cent (334) of the patients had no or slight pain. At the latest follow-up examination, 323 (87 per cent) of the patients (362 [87 per cent] of the hips) had no or slight pain in the limb over-all, and the average pain score was 40 points. The prevalence of pain in the thigh was between 5 and 10 per cent at all follow-up evaluations. At the most recent examination, twenty-two (6 per cent) of the patients (twenty-seven [6 per cent] of the hips) had pain in the thigh. All but one of them rated the pain as mild or moderate and stated that it did not limit activities. The remaining patient had a score of 8, indicating severe pain. Fifty patients (13 per cent) at the one and two-year follow-up examinations and eighty-three patients (22 per cent) at the latest follow-up examination had a moderate or severe limp, as defined by The American Academy of Orthopaedic Surgeons/Hip Society questionnaire39. This trend toward an increased limp and similar percentages were also observed with the arthroplasties done with cement and the hybrid arthroplasties performed at our institution, and they have generally been considered signs of polyarticular disease or general debility, or both. As previously stated, no attempt was made to classify patients by Charnley functional class12.
Radiographic Results Postoperatively, the calcar of the femur was in contact with the collar of the prosthesis in 329 hips (79 per cent). The femoral stem was aligned in neutral in 395 femora (95 per cent), in varus in eight (2 per cent), and in valgus in thirteen (3 per cent). Metaphyseal fit was good in 366 femora (88 per cent) and poor in fifty (12 per cent). Isthmal fill was good in 329 femora (79 per cent) and poor in eighty-seven (21 per cent). The following radiographic findings were noted on comparison of the serial radiographs, including those made at the latest follow-up examination. Calcar rounding was present in 259 femora (62 per cent) and was generally non-progressive after six months postoperatively. Resorptive bone-remodeling was seen on the radiographs of 207 femora (50 per cent). Resorption was limited to mottling in zone 1 or zone 7 in 160 femora (77 per cent; 38 per cent over-all). The most common pattern of mottling was in zones 1 and 7 (121 femora; 29 per cent over-all), followed by zone 7 only (seventy-five femora; 18 per cent over-all). Forty-seven femora (11 per cent over-all) had a combination of mottling and cortical thinning. This loss of cortical density, or cortical thinning, was generally limited to the proximal zones and occurred most commonly in zones 2 and 6 (seventeen femora; 4 per cent) (Figs. 3-A, 3-B, and 3C). Only ten femora (2 per cent) had cortical thinning distal to zones 2 and 6. No femur had a complete loss of the distal cortex.
Only twenty-four femora (6 per cent) had endosteal spot welds on two sides of the implant—that is, zones 2 and 6 of Gruen et al.26 or zones 9 and 13 of Johnston et al.39—and forty-two (10 per cent) had them on only one side. Some degree of distal cortical hypertrophy occurred in nearly half (204) of the femora. New periosteal bone was seen in fifty-four femora (13 per cent); endosteal bone, in fifty-eight (14 per cent); and combined endosteal and periosteal bone, in ninety-two (22 per cent). The most common location of periosteal cortical hypertrophy was zones 3 and 5 (seventy-five femora; 18 per cent), followed by zone 5 alone (fifty-eight femora; 14 per cent). The most common location of endosteal cortical hypertrophy was zones 3 and 5 (123 femora; 30 per cent), followed by zone 3 alone (fifteen femora; 4 per cent) or zone 5 alone (twelve femora; 3 per cent). The hypertrophy ranged from a small amount of dense bone to an exaggerated response that encased the stem and ballooned the cortex (Figs. 4-A, 4-B, and 4-C).
Over-all, the most commonly observed pattern of bone-remodeling was no changes whatsoever, followed by resorption in zone 1 or zone 7, or both, and hypertrophy in zone 3 or zone 5, or both. Radiodense lines were extremely common around the smooth, distal part of the stem, forming a halo pedestal in 166 femora (40 per cent). A shelf pedestal was seen in thirty-seven femora (9 per cent). Seven hips had a small area of cystic radiolucency in the calcar of the femur, which was presumed to be due to osteolysis. These lesions were all less than one centimeter in diameter and did not progress markedly on the serial radiographs. No distal osteolysis was seen. There was no shedding of the porous coating or breakage of the stem. Forty-four stems (11 per cent) had measurable subsidence, which was more than five millimeters in five stems (1 per cent). Serial standard anteroposterior radiographs of the pelvis were analyzed for evidence of continued subsidence. The subsidence of all but two stems stabilized within six months postoperatively. One of the other two stems stabilized within a year. The second stem had progressive subsidence and shifted into varus angulation. It was the only stem in the series with a circumferential radiodense or radiolucent line and the only stem that was considered unstable at the latest follow-up examination. This patient had severe rheumatoid arthritis and poor bone quality. She had no pain in the hip after the total hip arthroplasty and died three years postoperatively. Twenty-one non-displaced, intraoperative longitudinal cracks of the proximal-medial portion of the femur were fixed with cerclage wire. There were also two fractures at the level of the tip of the implant: one was treated with internal fixation and the other, with a long-stem implant. Thus, the prevalence of fracture was 6 per cent (twenty-three femora).
Statistical Analysis (Table I)
The ability to achieve a good metaphyseal fit was significantly decreased by poor trabecular bone (as determined by the trabecular index) (p = 0.05), and isthmal fill was significantly decreased by poor trabecular and poor cortical bone (p < 0.001 for both) (as determined with the trabecular and cortical indices). Proximal bone resorption and distal cortical hypertrophy were both significantly more common in patients who had a lower trabecular index (p < 0.001) or cortical index (p < 0.001). Endosteal spot welds were significantly more common in femora with a poor trabecular index (p = 0.03). Subsidence of the femoral component was significantly associated with a lack of contact of the collar of the prosthesis with the calcar of the femur (p = 0.01), a lower cortical index (p = 0.04), and a lack of metaphyseal fit (p < 0.001). Taken together, the data support the observation that poor bone quality predisposes to poor fit and fill and eventual proximal resorptive bone-remodeling and distal cortical hypertrophy. Conversely, femora with good bone quality tended to have little or no remodeling after insertion of the stem. However, these variables and radiographic findings were not predictive of the clinical outcome in the case of this particular femoral component.
The original purpose of this study was to determine postoperative radiographic features that were predictive of the ultimate failure or success of total hip arthroplasty performed with insertion of a Mallory-Head femoral component without cement. Many radiographic signs that seemed peculiar to this stem were observed early in the series, without the knowledge of which of them might be important in the future. As the patients with these stems were followed, it became obvious that the clinical results remained excellent and certain radiographic patterns persisted. Thus, our purpose became more to describe the results and to illustrate the unique radiographic changes that occur after insertion of this stem. These findings are important for two reasons. First, the continued use of implants without cement must be based, at least in part, on existing clinical results. Until now, there have been very few reports on the use of a titanium tapered femoral stem inserted without cement. The excellent short-term clinical and radiographic results in the present series support the continued use of this type of design for fixation without cement. Second, it is important to recognize the radiographic parameters that are associated with stability or instability of a given femoral stem. The requirements for stable fixation of porous-coated implants have been studied extensively, and many principles are well accepted. Initial stability of the implant is critical to minimize micromotion and to allow bone ingrowth7,57,70. The surgeon must be able to fit the implant in the metaphysis and to fill the isthmus in order to achieve stability19,20,43,52,66. The porous surface must be biocompatible, resist abrasion and corrosion, and form a strong bond with the implant substrate. The optimum pore size has been determined to be 100 to 400 micrometers7. However, many issues concerning fixation of stems without cement remain controversial. The choice of implant material as well as the extent of the porous coating and the method of its application are still debated. Support can be found for the use of custom or modular stems59, for one or many different anatomic-type stems10,55, and for straight stems5,6,47,48,78. Tapering of the stem has its proponents48,78, but it is associated with a risk of the femur splitting during insertion or during normal activities75. The desirability of a collar for stability is also debated7,76. Bone ingrowth-enhancing factors applied to the porous surface have shown promising early results, but their use remains controversial14,25. In fact, the need for a porous surface is debatable as non-porous stems, with or without ridges or texturing, have demonstrated excellent stability, bone apposition, and clinical results46,56,78. Use of early designs of femoral components without cement had generally good results in the short term8,33. However, as more and longer-term studies8,11,15,27,33,34,41,51 have become available, the results have often neither met expectations nor rivaled the results of total hip arthroplasty performed with use of newer cementing techniques. Kim and Kim reported failure of 12 per cent (fourteen) of 116 porous-coated anatomic stems at a minimum of six years43. Heekin et al. reported instability of five of 100 stems that had been followed for five to seven years34, and Maric and Karpman reported a 10 per cent rate of subsidence and an 8 per cent rate of revision of fifty-two porous-coated anatomic stems after an average of only 2.4 years51. Pain in the thigh has been a persistent problem associated with the porous-coated anatomic implant, with reported rates of 4 to 34 per cent, and is occasionally severe enough to warrant revision of a well fixed stem8,11,15,27,33,34,36,41,51,60. Shedding of the porous coating from more than 20 per cent of porous-coated anatomic stems has been demonstrated, and osteolysis has been seen around as many as one-third of porous-coated anatomic stems8,11,43,51. Maloney and Harris reported loosening of five of twenty-five Harris-Galante stems that had been followed for a minimum of two years49; the average clinical scores were much less than those after comparable hybrid total hip arthroplasties. Kim and Kim reported failure of 10 per cent (eight) of eighty-two Harris-Galante stems after an average of sixty-two months, with a persistent decline in results during the follow-up period42. Martell et al. reported eleven unstable stems, four of which were revised, an average of sixty-seven months after 121 Harris-Galante total hip arthroplasties52. Thus, initial and late instability of the Harris-Galante femoral component has been reported. Osteolysis has also become a major concern with regard to this stem, occurring in more than 10 per cent of hips and often compromising the stability of the implant and the strength of the femoral bone50,52,74,77. Delamination of the porous coating has also been seen52. The anatomic medullary locking femoral stem inserted without cement has been associated with long-term clinical and radiographic success. Engh and Massin reported an 88 per cent rate of survival, as defined by stable fixation, for 343 stems at eight years postoperatively19. A 1 per cent rate of revision of 959 stems that were followed for two to twelve years, with a ten-year rate of survival of 96 per cent, was also reported21. With the current techniques and implant design, stable fixation has been achieved in more than 90 per cent of hips21. Shaw et al. confirmed these results in a short-term study, reporting a 98 per cent rate of satisfaction at an average of forty months after 178 arthroplasties with an anatomic medullary locking total hip prosthesis66. However, pain in the thigh has been reported with use of the anatomic medullary locking stem, occurring in as many as 34 per cent of patients20,27,28,66. Also of concern is the sometimes extensive bone resorption associated with use of the anatomic medullary locking stem. Engh et al. reported moderate or severe resorption, as determined by the degree of bone loss seen radiographically, around 12 per cent of 307 stems20. A study of retrieved specimens demonstrated regional reductions of bone mineral content of 7 to 78 per cent and found that bone density never returned to normal after insertion of a stem24. Dual-energy x-ray absorptiometry revealed an over-all decrease in bone mineral content of 7 to 52 per cent around anatomic medullary locking stems in ten cadaveric femora after seventeen to eighty-four months in situ23. Partly because of these early shortcomings and concerns and partly because of the number of design variables possible, a tremendous variety of total hip prostheses designed to be inserted without cement have become available to orthopaedic surgeons. However, many design features seem to have come about with little scientific reasoning or clinical experience to justify them. It remains uncertain if enough experience can be garnered with one design to obtain meaningful data. We have been fortunate in that we have had a great deal of experience with one total hip system performed without cement. Our interpretations of the clinical and radiographic findings reported here are based on our experience with the Mallory-Head femoral component. Most of these interpretations are purely speculative, as no retrieval or revision operations of well fixed or aseptically unstable stems were performed. It must be emphasized that these results are also not conclusive, as much longer follow-up is necessary to determine long-term stability and longevity. Thus far, use of the Mallory-Head component has yielded excellent clinical results as measured by the low prevalence of pain in the lower limb over-all and in the thigh and the lack of revisions for instability. These results compare favorably with our results with cemented stems as well as with those of other authors who used newer cementing techniques30,31,54,62 and are a marked improvement compared with our experience with the porous-coated anatomic femoral stem (Howmedica, Rutherford, New Jersey)6. The explanations for the short-term clinical superiority of this stem are poorly understood. We believe that the tapered stem creates a wedging effect, which enhances early and late stability. The circumferential titanium plasma-spray appears to provide a stable interface that promotes bone ingrowth and prevents distal osteolysis. Pain in the thigh is infrequently associated with the Mallory-Head femoral component. The potential causes of pain in the thigh were the subject of a previous review60. Two causes, instability of the stem and a tight distal fill with a rigid stem, are avoided with the Mallory-Head stem. The statistical analysis in the present study confirmed the findings of an earlier study6 that pain in the thigh was not associated with any radiographic variable, such as initial subsidence, a shelf pedestal, or a halo pedestal (indicating micromotion). The lack of major resorptive bone-remodeling in this series is probably the result of the use of a titanium alloy stem with porous coating only on its proximal part. Titanium is 50 per cent less stiff than chromium-cobalt, thereby increasing the proximal transfer of stress3,35,58,71,72. In experimental studies and in studies on canine models, titanium stems produced much less bone resorption than equivalent chromium-cobalt stems, as confirmed histologically and with dual-energy x-ray absorptiometry scans4,25. Limiting the extent of porous coating has also been shown to be a factor in the reduction of stress-shielding4,18. Pre-existing osteopenia and the resultant need for a stem with a large diameter is a major, if not the most critical, factor in the production of resorptive bone-remodeling. However, only severe osteopenia led to resorptive bone-remodeling distal to the very proximal aspect of the femur (zones 1 and 7) after insertion of this stem. While cortical hypertrophy has been reported to occur around as many as 30 per cent of Harris-Galante stems, 19 per cent of anatomic medullary locking stems, and 14 per cent of porous-coated anatomic stems28,33,37,61, it has not been reported to the frequency or extent with which it was found in the present series. We were surprised and intrigued by this finding. The cortical hypertrophy in our series generally developed within six months after insertion of the stem and changed little after that time. The potential causes of this distal cortical hypertrophy are poorly understood. Both the decreased rigidity and the tapering of the stem may create a proximal-to-distal stress transfer, thus loading the femur at the distal aspect of the stem. The cortical hypertrophy may also be due to a lateral bending moment created by insertion of a straight stem into a curved femur. Whatever its cause, cortical hypertrophy is common and occasionally occurs to an exaggerated degree after insertion of a Mallory-Head stem. Osteolysis has become the major concern after insertion of stems without cement. It can occur in association with both stable and unstable femoral components and can lead to instability of the stem and diaphyseal weakening50,74,77. Osteolysis has been shown to result from polyethylene debris tracking into the effective joint space around non-circumferentially coated stems64 or from the porous coating itself if it delaminates or sheds. Femoral endosteal lysis seems to occur earlier and to progress more rapidly around stems inserted without cement than around those inserted with cement74. In the present series, early wear of the polyethylene and pelvic osteolysis were becoming important problems. However, the circumferential plasma-spray on this stem apparently protects against the distal progression of debris, as no distal femoral lysis has been seen to date. The manufacturer of this stem reports a bond strength between the plasma spray and the substrate of 6000 pounds per square inch (41,369 kilopascals), which protects against shedding of the porous particles. We observed no shedding of the porous coating in this series, confirming the findings of an earlier report on more than 900 total hip arthroplasties47. The lack of shedding of the porous coating and of distal osteolysis are encouraging in this short-term review. However, fretting and wear of titanium alloy with release of extensive debris into the tissues has been reported with other titanium hip implants1,38,45. Titanium has been shown to cause an exaggerated inflammatory response32. Therefore, although distal osteolysis around the stem has not been noted to date, much longer follow-up is necessary to exclude this as a complication. The fact that no stems broke in this and another series5 is related to the plasma-spray process, which maintains a fatigue strength of the titanium alloy of 70,000 pounds per square inch (482,633 kilopascals), a major improvement compared with sintering and diffusion bonding techniques. Calcar rounding, a common finding in this series, was thought to be due to maturation of the osteotomy site in the femoral neck. A halo pedestal was considered a sign of micromotion and a shelf pedestal, a sign of endosteal stimulation at the tip of the implant. The shelf pedestal was not associated with other radiographic signs of instability, as previous authors have found22. None of these signs was considered clinically significant, as confirmed with statistical analysis. Engh et al. described the radiographic signs of fixation and stability of chromium-cobalt stems inserted without cement22. The lack of reactive lines and the presence of spot welds around the porous surface were considered major signs of osseous integration; the absence of subsidence was considered a minor sign of stability22. In their comprehensive review, spot welds and proximal resorption were found in most hips that had stable fixation, and the lack of spot welds was a hallmark of unstable fixation. A pedestal was important if it was associated with an unstable stem tip. In the present series, endosteal spot welds were distinctly uncommon, despite apparently stable fixation. The lack of spot welds may be a reflection of the lack of resorption of surrounding bone, leaving indistinguishable areas of bone ingrowth. This phenomenon was suggested by Engh et al. in the evaluation of smaller-diameter stems with stiffnesses approaching that of the femur22. The initial subsidence observed in forty-four hips (11 per cent) in the present series might predict a lack of ingrowth and an eventually unstable stem. However, only one stem continued to subside, and it shifted into varus angulation after a six to twelve-month period. No other stem subsided to a measurable degree from twelve months postoperatively to the most recent follow-up examination at an average of 3.7 years. Eventual stability after limited subsidence may be the natural course of a tapered stem without a functional collar. Therefore, it has been our experience that endosteal spot welds, resorptive bone-remodeling, and a lack of early subsidence are not reliable radiographic indicators of the stability of this tapered titanium stem. In addition, distal cortical hypertrophy and formation of a small pedestal appear to be benign and coincidental radiographic findings. Instability of the stem should be judged only by progressive subsidence and the presence of radiolucent or radiodense lines at the porous surface. These radiographic findings should be kept in mind when this titanium stem is evaluated, and the criteria for stability of the stem should be modified accordingly. In other words, the criteria set forth by Engh et al.22 for stability of chromium-cobalt femoral stems are not appropriate for the Mallory-Head femoral component. This study has several limitations. As with all radiographic reviews, it is dependent on the observations and interpretations of the reviewers; there were no retrieval studies or findings at revision to support the radiographic interpretations. Bone ingrowth has been demonstrated in titanium implants and plasma-sprayed surfaces46,72, but we cannot draw conclusions regarding the osseous integration of these stems on the basis of this review. Sensitive means for measuring bone-remodeling, such as dual-energy x-ray absorptiometry scanning, were not used. It is well accepted that a minimum 30 per cent loss of bone mineral content is necessary before osteopenia is evident radiographically. Therefore, the present study can provide only a crude judgment of bone-remodeling changes. Lastly, this is a short-term review of one particular titanium-alloy femoral component without comparison with a matched or randomized group in which a femoral stem was inserted with cement. Until long-term studies document the efficacy of femoral fixation without cement, cement should be considered the preferred method of fixation for femoral components30,31,54,62,65. We do not suggest that the Mallory-Head stem should replace or even rival stems inserted with cement but that it should be an alternative femoral component for surgeons who want to use fixation without cement.
NOTE: The authors thank Robert Hardie for assistance with the statistical analysis.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  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. |