- Email: [email protected]
Hydroxyapatite-coated femoral components in total knee arthroplasty
The Journal of Arthroplasty Vol. 18 No. 7 2003
Hydroxyapatite-Coated Femoral Components in Total Knee Arthroplasty Medium Term Results Aradhyula Narasimha Murty, FRCS, G. Scott, FRCS, and M. A. R. Freeman, FRCS
Abstract: Thirty-six consecutive total knee arthroplasties with hydroxyapatite (HA)-coated femoral components were prospectively followed up according to a standardised protocol for a mean period of 10 years (range, 7–11 years). Clinical and radiologic examination was performed at each follow-up evaluation. One patient (1 knee) was lost to follow up at 7 years, and 4 other patients (6 knees) died. Radiographs were studied for signs of loosening and presence of lucencies. Survival analysis was performed using a life table with revision as the criterion for failure. Two knees (1 patient) were revised because of aseptic loosening. The tibial component of one other knee was revised because of osteolysis in the medial tibial condyle. In the remaining knees, no radiologic evidence of loosening was seen. The survival of the femoral component at a mean follow-up time of 7 years, when 31.5 knees were at risk, was 94% with revision for femoral aseptic loosening as the end point. At 10 years, when 24 knees were at risk, the survival remained 94%. Key words: hydroxyapatite– coated femoral component, total knee arthroplasty. © 2003 Elsevier Inc. All rights reserved.
Cemented total knee arthroplasty has been shown to give consistent results over 10 years, with a survival rate of more than 95% [1,2] Cemented knee arthroplasty is associated with loss of bone stock in cases of aseptic loosening or infection, causing problems with revision [3,4] and a possible adverse effect of cement on leukocyte function [5]. To overcome these and other past problems, uncemented designs have been introduced. To enhance uncemented fixation, porous or hydroxyapatite (HA) coatings have been added. In animal experi-
ments [6,7] and in retrieval studies on hips [8 –10], HA-induced bone ingrowth into gaps has been shown. Encouraging results have also been reported from hip arthroplasties using HA-coated femoral prostheses [11–13]. Radiostereophotogrammetric analysis (RSA) on tibial components coated with HA have shown that comparable stability is obtained to that achieved by cementing at 2 to 5 years after replacement [14,15]. However, we are not aware of any studies that have reported the results of HA coating of the femoral component in total knee arthroplasty in the medium term. We hypothesized that stability provided by HA coating of the femoral components in total knee arthroplasty would be comparable to that achieved by cemented fixation. We began an observational study in 1991 to examine the radiologic and clinical outcomes. We report on survival of these HAcoated femoral components at a mean follow-up time of 10 years (range, 9 –11 years).
From the Bone and Joint Research Unit, The Royal London Hospital, Whitechapel, London, United Kingdom. Submitted June 29, 2001; accepted May 23, 2003. No benefits or funds were received in support of this study. Reprint requests: Aradhyula Narasimha Murty, FRCS, c/o John Ireland, FRCS, 17 Kings Avenue, Woodford Green, Essex, UK IG8 OJD. © 2003 Elsevier Inc. All rights reserved. 0883-5403/03/1807-0006$30.00/0 doi:10.1054/S0883-5403(03)00324-3
844
Hydroxyapatite-Coated Femoral Components in TKA • Murty et al.
845
Fig. 1. Anterior and posterior oblique views of one of the explanted F/S modular total knee arthroplasties with HA coating under standard lighting conditions.
Materials and Methods The study group consisted of 36 knees in 30 patients in the United Kingdom. These consecutive patients with disabling knee arthritis underwent total knee arthroplasty using Freeman Samuelson (F/S) modular prostheses (Sulzer Orthopaedics, Mill Lane, Alton, and Finsbury Instruments, Leatherhead, UK) with HA-coated femoral components. In 16 of these, both femoral and tibial components were HA coated. In the remaining components, only the femoral component was HA coated, and the tibial component was secured with proximally placed cement. Infected cases were excluded from the study. Design The design of the prosthesis has been described before [16]. To summarize, the femoral component is manufactured from polished cobalt-chrome alloy with a constant radius of 24 mm for the tibiofemoral articulation. The component has a trochlear groove with an independent radius for patella tracking. The component is supplied for fixation with either 2 condylar pegs or a short stem, according to surgeon preference. The internal surface is grit-blasted to produce a surface roughness of 3 Ra to which HA coating was added. The tibial component is made of titanium alloy with 2 pegs and an
80-mm stem. The undersurface of the tibial component was HA coated in some cases. The UHMWPE insert was located in the tibial tray. This had the same radius as the femoral component but was sufficiently shallow to allow subluxation at an extreme load. There was no central stabilizer (Fig. 1). Hydroxyapatite Coating Hydroxyapatite coating was applied by plasma spray (Plasma Biotal, Tideswell, UK) to the shotblasted surface of the prosthesis, excluding the condylar pegs or the stem. Its characteristics were purity, ⬎98% HA; crystallinity, ⬎75%; shear strength, 20 to 40 Mpa; and thickness, 80 to 120 m. The surgical technique has been described before [16]. The 2 senior authors (M.A.R.F., G.S.) or trainees under their supervision conducted all the surgeries. Each patient received prophylactic antibiotics, which generally consisted of 3 doses of flucloxacillin. After the first 24 hours, the patients were allowed to bear weight as tolerated using 2 crutches. The crutches were retained for 6 weeks. Follow-Up Evaluations The study cases were followed up prospectively according to a standardized protocol at 6 months and 1, 2, 3, 5, 7, 8, 9,10, and 11 years. At each
846 The Journal of Arthroplasty Vol. 18 No. 7 October 2003
Fig. 2. Nomenclature of different zones of femoral interface.
follow-up evaluation, functional and clinical status were recorded, and weight-bearing anteroposterior and lateral radiographs taken. Twenty-seven of the 36 knees in this series have been regularly reviewed clinically or over the telephone and radiographed as described previously. One further patient (2 knees) remains under review at 11 years but has declined radiography since year 5. Four knees in 2 patients were followed up until the patients died at 7 and 8 years. They are included in the review until then. One patient (1 knee) emigrated at 7 years, and was reviewed until then. This patient was categorized as lost to follow-up at 7 years. These 5 knees were functioning well when last seen. Two patients (2 knees) had radiographs only at 1 year because of severe comorbidity. Both later died at 4 and 8 years. The former knee was not revised at 4 years, but the functional status is not known. The latter knee was functioning satisfactorily on the basis of telephone review at 8 years. These patients were included in the review until their deaths. Radiologic Evaluation The follow-up radiographs were evaluated for alignment of the components to the tibia and femur, overall alignment, and femoral component loosening. The alignment was studied as per criteria described by the American Knee Society [17]. In the lateral view, the femoral interface was divided into different zones (Fig. 2). In the anteroposterior view, lucencies and osteolysis were investigated around the stem when present. The following radiologic signs were taken as signs of loosening. These are based on an earlier study conducted in the same unit on uncemented knee arthroplasties [18]:
i) A radiolucent line more than 2 mm in thickness. ii) Progressive radiolucent lines. The lines were considered progressive if the thickness increased or if it progressively involved the adjacent zones. iii) Proximal migration of femoral component manifest by medial and lateral bone protrusions in the anteroposterior view or increased distance between the posterior lip of the femoral prosthesis and the tip of posterior condyle in the lateral view. iv) More than 3° change in the position of the femoral component in the anteroposterior or lateral view. The results were compared with a previously published report of cemented and cementless (nonHA) fixation of F/S components. These patients also underwent surgery performed by the senior authors using the same surgical technique for the same clinical indications. These procedures differed only in the manner of femoral component fixation [18]. Survival Analysis The life table method was used with revision for aseptic loosening of the femoral component as the criterion for failure to calculate the survival analysis.
Results Thirty-six knees comprised the study group. The demographic data of the group is summarized in Table 1. The average follow-up time was 10 years (standard deviation, 0.83; range, 7–11 years).
Table 1. Demographic Data Age Gender ratio, M:F Side, R:L Diagnosis Primary OA RA Psoriatic arthritis Post-traumatic arthritis Avascular necrosis Previous surgery Patellectomy Arthrodesis Synovectomy Tibial tubercle advancement
Average, 63 y; SD, 9.84 18:18 18:18 30 2 1 1 2 2 1 2 1
Abbreviations: M, male; F, female; R, right side; L, left side; OA, osteoarthritis; RA, rheumatoid arthritis.
Hydroxyapatite-Coated Femoral Components in TKA • Murty et al.
847
Survival Analysis Over the period of the study, 3 revisions have occurred. Two of these were revision of the femoral and the tibial components and one was of the tibial component only. Two cases of aseptic femoral loosening occurred in one patient with bilateral simultaneous arthroplasty. The right knee failed at 1 year and was revised. The other knee failed at 7 years. Both the tibial components were implanted with 5° varus tilt. The femoral components had a reciprocal valgus tilt. At revision of the right side, the femoral component was loose and a complex defect of the medial femoral condyle filled with soft tissue was noted. The tibial component was well fixed with bone growth on the undersurface of the tibial component. which was HA coated. At revision of the left side, the femoral component was again loose and had subsided on the femur, causing erosion of the distal condyles. The UHMWPE insert was partially delaminated. The tibial component in this case was also solidly fixed, and removal was extremely difficult. The third knee developed severe osteolysis of the medial tibial condyle without loosening of the HA-coated tibial component. The UHMWPE insert, which was only 6 mm thick, was found to have severe wear. The femoral component was found stable and was not revised. Therefore, it remains in this review. Histologic analysis showed no evidence of HA particles in the debris, but an abundance of polyethylene particles. Presumably, the osteolysis was caused by accumulated wear debris. The survival rate of HA-coated femoral components at a mean of 7 years after arthroplasty, when 31.5 knees were at risk, was 94% (95% confidence interval, 85.4 –100). At 10 years, the survivorship with revision for any reason is 90.5% (95% confidence interval, 78.5–100). However, the survival of the femoral component, with revision for loosening as the end point, was 94% (95% confidence interval, 84 –100). The survival curve is shown in Fig. 3, and the life table is shown in Table 2.
Fig. 3. The survival curve.
instability. At the final follow-up examination, no objectively demonstrable instability was seen. The patient had previously undergone patellectomy and tibial tubercle advancement. She also had undergone decompression for spinal canal stenosis. No evidence of loosening was seen on radiographs. We cannot explain the instability or the pain. Multiple surgeries causing proprioceptive dysfunction may play a role. The other patient (2 knees) was also symptomatic in the cervical and lumbar spine. Radiographs in this case did not show any evidence of loosening. The pain in this case, too, remains unexplained. One other patient had subjective instability. That patient had undergone no previous surgeries, and the symptoms remain unexplained. Activity Level All our patients were either retired or homemakers. No patient was involved in recreational sports. Uninterrupted Walking Duration Preoperatively, 65% of our patients had a maximum uninterrupted walking duration of more than 10 minutes, of which 7% could walk for more than 30 minutes. At final follow-up evaluation, these figures were 96% and 76%, respectively. The one patient who could not walk had multiple joint involvement because of rheumatoid arthritis.
Clinical Results
Use of Aids for Walking
The clinical results are summarized in Table 3. We did not use a clinical scoring system. We report on the pain level and present the preoperative and postoperative activity level in 7 activities of daily living, namely uninterrupted walking duration, use of aids for walking, ability to get out of a dining chair, ascending and descending stairs, bathing independently and conducting housework. Two patients (3 knees) experienced pain requiring analgesia. One of these patients had associated subjective
In our study, 17% of patients were using either crutches or a frame for walking before the surgery. At final follow-up evaluation, 60% of our patients were walking without the use of any aids, and 36% were using one stick. One patient was using a frame to walk. Ability to Rise from a Chair Preoperatively, 39% of our patients did not report any difficulty in rising from a dining chair,
848 The Journal of Arthroplasty Vol. 18 No. 7 October 2003 Table 2. Survival Analysis (Life Table) Years Since Surgery
Start (n)
Failed (n)
Died (n)
Lost (n)
At Risk (n)
Survival Rate
Confidence Limits
0–1 1–2 2–3 3–4 4–5 5–6 6–7 7–8 8–9 9–10
36 36 35 35 34 34 32 30 27 24
0 1 0 0 0 0 1 0 0 0
0 0 0 1 0 2 0 2 1 0
0 0 0 0 0 0 1 0 0 0
36 36 35 34.5 34 33 31.5 28.5 25.5 20
100 97.2 97.2 97.2 97.2 97.2 94 94 94 94
1.0–1.0 1.0–1.0 .918–1.0 .918–1.0 .918–1.0 .918–1.0 .915–1.0 .856–1.0 .857–1.0 .84–1.0
54% could do it only by pushing off with their hands, and 7% were unable to stand up without help. At final follow-up examination, 63% could perform this activity with ease and 37% could perform this activity with difficulty using a push off.
some help, and 15% either used a shower or were unable to bathe themselves. At final follow-up evaluation, these figures had improved to 60%, 30%, and 10%, respectively. Housework
Ascending and Descending Stairs Before surgery, 86% of our patients could negotiate stairs using a reciprocating gait provided they used a hand rail or climbed stairs with a nonreciprocating gait; 14% were unable to negotiate stairs. At final follow-up evaluation, 84% were climbing and descending stairs using a reciprocating gait .Only 10% needed to use hand rails or had a nonreciprocating gait. One patient was unable to climb stairs, and one patient did not have stairs at home. Bathing Before the knee arthroplasty, 33% of the cohort could bathe without help, 52% could bathe with
Table 3. Clinical Results Pain-free/Occasional analgesia Range of motion
Alignment Instability Stair climbing with reciprocal gait Uninterrupted walking duration
91% FFD Flexion
2 patients (⬍10°) Average, 100° SD, 14° Range, 70°–120° Extensor lag 0 Average 7.2° Valgus SD 1.13° Range 0–10° 2 patients (subjective) 84%
Preoperatively, 11% of our patients were able to do all their housework, 63% were doing part of their housework, and 26% were either unable or had not attempted. At the last follow-up evaluation, the numbers were 63%, 27%, and 10%, respectively. Radiologic Results The mean alignment of the femoral component was 6.9° valgus ⫾ 1.7° (␣ ⫽ 96.9°). The mean alignment of the tibial component was 0.6° varus ⫾ 1.4°. The femoral component was flexed by a mean of 3.8° ⫾ 1.4° (␥ ⫽ 3.8°). The slope of the tibial component was directed posteriorly by a mean of 1° ⫾ 2.1° (␦ ⫽ 89°).The mean tibiofemoral alignment was 7.2° ⫾ 1.1° valgus. For 32 knees, we had a radiograph from at least a 5-year follow up. Of these, 12 knees had no identifiable radiolucent lines. In the remaining 20 knees, 30 radiolucent lines were seen (Fig. 4). None of these were progressive, more than 2 mm in thickness, or associated with other signs of prosthetic loosening. On the basis of the criteria described, no loose femoral components were noted. However, because the radiographs taken were not under fluoroscopic control, some radiolucent lines may have been hidden (Fig. 5).
⬎30 min
76%
Discussion
10–30 min ⬍10 min
20% 4%
This study reports 94% survival of the HA-coated femoral component in total knee arthroplasty at an
Hydroxyapatite-Coated Femoral Components in TKA • Murty et al.
Fig. 4. Distribution of radiolucent lines in different zones.
average follow-up time of 10 years, with femoral revision for aseptic loosening as the end point. Although these results are encouraging, they are not as good as those for cemented knee arthroplasties. We have concentrated this report on the survival of the femoral components only because only 16 patients had HA coating for both components. From this subgroup, 3 tibial components have been revised, but none were found to be loose at revision. Two revisions were performed in 1 patient when the femoral component was found to be loose, but the tibial component was solidly fixed with bone growth on the undersurface of the tibial component. The third revision was for osteolysis of the medial tibial condyle. The tibial component was not loose and was extremely difficult to remove. In this small experimental study, not all component sizes were available. The allocation of the femoral component (whether pegged or stemmed) was determined by availability. Similarly, the correct tibial size was not always available with HA coating. Therefore, some patients received a cemented tibial component. Aseptic loosening of the femoral component is generally believed to be less frequent than that of the tibial component in total knee arthroplasty [19]. The failure rate of uncemented femoral components was been reported to be between 0% and 5% in some short- and long-term follow-up reports [3,20,21]. However, other studies have reported up to 15% failure at 5 years [22]. A previous study on uncemented non–HA-coated F/S femoral components conducted in this unit reported 10% failure rate at 7 years [18]. Thus, the survival of F/S HA-coated femoral components seems to be better than that of most other uncemented prostheses in the medium term. The stability of any prosthesis depends on the interlock achieved between the bone and the im-
849
plant surface [23]. Researchers have shown that the cuts made by a saw are not uniformly smooth [24], and gaps are inevitable between the implant and the bone surface. Cementing under pressure fills these gaps and provides the necessary interlock, which can resist shear and rotary stresses [23]. HA can induce chemical bonding, initially providing the necessary interlock [10]. HA can also conduct bone growth onto the surface of the implant across a gap up to 5 mm wide [6,7]. It has been shown to enhance stability by converting fibrous tissue to bone under load [25]. Animal experiments have shown that in a cancellous site HA-conducted bone growth to the implant occurs in 4 to 6 weeks [26]. Provision of initial stability to enable the osteoconductive properties of HA to take effect seems to be important; micromotion of 150 m has been shown to inhibit bone formation [27]. In our series, one of the 2 knees that failed did so early, lending credence to this hypothesis. The other knee in the same patient failed after 7 years .The alignment of the prostheses in these 2 knees was poor, producing an oblique joint line and presumably abnormal interface loading. This may have lead to failure, whatever the method of fixation. However, even in these knees, both components showed evidence of bone ongrowth (Fig. 6). A reported complication of HA coating at the hip is the generation of particulate matter leading to third body wear. HA particles found in osteolytic areas are associated with an inflammatory reaction [28]. One of our knees had severe osteolysis in the tibial condyle, which required tibial revision. The tibial and femoral components were well fixed. The UHMWPE insert, which was only 6 mm in
Fig. 5. Anteroposterior and lateral radiographs of a typical patient at 6 years after surgery.
850 The Journal of Arthroplasty Vol. 18 No. 7 October 2003
Fig. 6. (A) An ultraviolet photograph of the internal surface of the femoral component. The U-V light has fluoresced the bone and appears bright white. The light grey areas are the HA coating without bone ongrowth and the dark grey area show where the HA coating is absent. (B) An ultraviolet photograph of the undersurface of the tibia tray of the component. As in panel A, bright white areas are ongrown bone; light grey areas are exposed HA without ongrowth, and dark gray areas are where the HA is absent.
initial thickness, showed severe wear. On microscopy of the contents of the osteolytic cavity, polythene debris was abundant but HA particles were not detected. Unfortunately, the polythene component was not further examined, so that we have no definite explanation for severe wear in this case. Because the femoral component was well fixed and was not revised, this knee did not represent a failure as defined here. HA did not seem to be directly responsible for osteolysis, and HA fixation withstood a large polythene load. Radiostereophotogrammetric analysis (RSA) has shown that migration of HA-coated tibial components is comparable to that with cement over 2 to 5 years [14,29]. The stability with regard to micromotion and migration is superior to that of non–HAcoated implants [30]. In light of these findings, we believe that HA-coated TKA should be evaluated with guarded optimism because it has the potential for conserving bone stock. However, in our experience to date, cemented fixation continues to be the most reliable form of fixation.
2.
3.
4.
5.
6.
7.
8.
References
9.
1. Robertsson O, Scott G, Freeman MAR: Ten year survival of the cemented Freeman Samuelson pri-
10.
mary knee arthroplasty. J Bone Joint Surg Br 82:506, 2000 Scuderi GR, Insall JN, Windsor RE, Moran MC: Survivorship of cemented knee replacements. J Bone Joint Surg Br 71:798, 1989 Fanning JW, Joseph J Jr., Kaufman EE: Follow up on uncemented total knee arthroplasty. Orthopaedics 19:933, 1996 Hungerford DS, Krackow KA, Kenna KA: Cementless total knee replacement in patients 50 years old and under. Orthop Clin North Am 20:131, 1989 Petty W: The effect of methylmethacrylate on bacterial phagocytosis and killing of human polymorphonuclear leucocytes. J Bone Joint Surg Am 60:752, 1978 Soballe K, Hansen ES, Brockstedt-Rasmussen H, et al: Gap healing enhanced by hydroxyapatite coating in dogs. Clin Orthop 272:300, 1991 Stephenson PK, Freeman MAR, Revell PA, et al: The effect of hydroxyapatite coating upon the in growth of bone into cavities in an implant. J Arthroplasty 6:51, 1991 Bauer TW, Geesink RCT, Zimmerman R, McMahon JT: Hydroxyapatite coated femoral stems. J Bone Joint Surg Am 73:1439, 1991 Furlong RJ, Osborn JF: Fixation of hip prosthesis by hydroxyapatite ceramic coatings. J Bone Joint Surg Br 73:741, 1991 Geesink RGT, De Groot K, Klein CPAT: Bonding of
Hydroxyapatite-Coated Femoral Components in TKA • Murty et al.
11.
12. 13.
14.
15.
16.
17.
18.
19.
20.
bone to apatite coated implants. J Bone Joint Surg Br 70:17, 1998 Geesink RGT, Hoefnagels NHM: Six- year results of hydroxyapatite coated total hip replacement. J Bone Joint Surg Br 77:534, 1995 Kroon PO, Freeman MAR: Hydroxyapatite coating of hip prostheses. J Bone Joint Surg Br 74:518, 1992 McNally SA, Shepperd JAN, Mann CV, Walczak JP: The results at nine to twelve years of the use of a hydroxyapatite coated femoral stem. J Bone Joint Surg Br 82:378, 2000 Nillson KG, Karrholm J, Carlsson L, Dalen T: Hydroxyapatite coating versus cemented fixation of the tibial component in total knee arthroplasty. J Arthroplasty 14:9, 1999 Onsten I, Nordquist A, Carlsson AS, et al: Hydroxyapatite augmentation of the porous coating improves fixation of tibial components. J Bone Joint Surg Br 80:417, 1998 Freeman MAR, Samuelson KM, Bertin KC: Freeman-Samuelson total knee arthroplasty. Clin Orthop 192:46, 1985 Ewald FC: The knee society total knee arthroplasty roentgenographic evaluation and scoring system. Clin Orthop 248:9, 1989 Chokalingham S, Scott G: The outcome of cemented versus cementless fixation of a femoral component in TKR with the identification of radiological signs for the prediction of failure. Knee 7:233, 2000 Insall JN, Haas SBComplications of total knee arthroplasty. p. 891Surgery of the knee2 EdIn , Insall JN, Windsor RE, Scott WN, et al. New York, Churchill Livingstone, 1993 Moran CG, Pinder IM, Lees TA, Midwinter MJ: Sur-
21. 22.
23. 24.
25.
26.
27.
28.
29.
30.
851
vivorship analysis of uncemented porous-coated anatomic knee replacement. J Bone Joint Surg Br 73: 848, 1991 Whiteside L: Cementless total knee replacement. Clin Orthop 309:185, 1994 Campbell MD, Duffy GP, Trousdale RT: Femoral component failure in hybrid total knee arthroplasty. Clin Orthop 356:58, 1998 Freeman MAR: Hydroxyapatite catalyst or conjurer? Orthopaedics 21:959, 1998 Toksvig-Larsen S, Ryd L: Surface characteristics following tibial preparation during total knee arthroplasty. J Arthroplasty 9:63, 1994 Soballe K, Hansen ES, Brockstedt-Rasmussen H, Bunger C: Hydroxyapatite coating converts fibrous tissue to bone around loaded implants. J Bone Joint Surg Br 75:270, 1993 Stevenson S, Goldberg VM, Tisdel CL, Dean JC: Application of hydroxyapatite in total knee arthroplasty. p. 99. In Insall JN, (ed). Current concepts in primary and revision total knee arthroplasty New York, Lipincott-Raven, 1996 Soballe K: Hydroxyapatite ceramic coating for bone implant fixation: mechanical and histological studies in dogs. Acta Orthop Scand Suppl 255:1, 1993 Bloebaum RD, Beeks D, Dorr LD, et al: Complications with hydroxyapatite particulate separation in total hip arthroplasty. Clin Orthop 298:19, 1994 Toksvig-Larsen S, Ryd L, Jorn L P, et al: Hydroxyapatite enhanced tibial prosthetic fixation. Clin Orthop 370:192, 2000 Nelissen RGHH, Valstar ER, Rozing P: Effect of hydroxyapatite on micromotion of total knee prostheses. J Bone Joint Surg Am 80:1665, 1998
Hydroxyapatite-Coated Femoral Components in Total Knee Arthroplasty Medium Term Results Aradhyula Narasimha Murty, FRCS, G. Scott, FRCS, and M. A. R. Freeman, FRCS
Abstract: Thirty-six consecutive total knee arthroplasties with hydroxyapatite (HA)-coated femoral components were prospectively followed up according to a standardised protocol for a mean period of 10 years (range, 7–11 years). Clinical and radiologic examination was performed at each follow-up evaluation. One patient (1 knee) was lost to follow up at 7 years, and 4 other patients (6 knees) died. Radiographs were studied for signs of loosening and presence of lucencies. Survival analysis was performed using a life table with revision as the criterion for failure. Two knees (1 patient) were revised because of aseptic loosening. The tibial component of one other knee was revised because of osteolysis in the medial tibial condyle. In the remaining knees, no radiologic evidence of loosening was seen. The survival of the femoral component at a mean follow-up time of 7 years, when 31.5 knees were at risk, was 94% with revision for femoral aseptic loosening as the end point. At 10 years, when 24 knees were at risk, the survival remained 94%. Key words: hydroxyapatite– coated femoral component, total knee arthroplasty. © 2003 Elsevier Inc. All rights reserved.
Cemented total knee arthroplasty has been shown to give consistent results over 10 years, with a survival rate of more than 95% [1,2] Cemented knee arthroplasty is associated with loss of bone stock in cases of aseptic loosening or infection, causing problems with revision [3,4] and a possible adverse effect of cement on leukocyte function [5]. To overcome these and other past problems, uncemented designs have been introduced. To enhance uncemented fixation, porous or hydroxyapatite (HA) coatings have been added. In animal experi-
ments [6,7] and in retrieval studies on hips [8 –10], HA-induced bone ingrowth into gaps has been shown. Encouraging results have also been reported from hip arthroplasties using HA-coated femoral prostheses [11–13]. Radiostereophotogrammetric analysis (RSA) on tibial components coated with HA have shown that comparable stability is obtained to that achieved by cementing at 2 to 5 years after replacement [14,15]. However, we are not aware of any studies that have reported the results of HA coating of the femoral component in total knee arthroplasty in the medium term. We hypothesized that stability provided by HA coating of the femoral components in total knee arthroplasty would be comparable to that achieved by cemented fixation. We began an observational study in 1991 to examine the radiologic and clinical outcomes. We report on survival of these HAcoated femoral components at a mean follow-up time of 10 years (range, 9 –11 years).
From the Bone and Joint Research Unit, The Royal London Hospital, Whitechapel, London, United Kingdom. Submitted June 29, 2001; accepted May 23, 2003. No benefits or funds were received in support of this study. Reprint requests: Aradhyula Narasimha Murty, FRCS, c/o John Ireland, FRCS, 17 Kings Avenue, Woodford Green, Essex, UK IG8 OJD. © 2003 Elsevier Inc. All rights reserved. 0883-5403/03/1807-0006$30.00/0 doi:10.1054/S0883-5403(03)00324-3
844
Hydroxyapatite-Coated Femoral Components in TKA • Murty et al.
845
Fig. 1. Anterior and posterior oblique views of one of the explanted F/S modular total knee arthroplasties with HA coating under standard lighting conditions.
Materials and Methods The study group consisted of 36 knees in 30 patients in the United Kingdom. These consecutive patients with disabling knee arthritis underwent total knee arthroplasty using Freeman Samuelson (F/S) modular prostheses (Sulzer Orthopaedics, Mill Lane, Alton, and Finsbury Instruments, Leatherhead, UK) with HA-coated femoral components. In 16 of these, both femoral and tibial components were HA coated. In the remaining components, only the femoral component was HA coated, and the tibial component was secured with proximally placed cement. Infected cases were excluded from the study. Design The design of the prosthesis has been described before [16]. To summarize, the femoral component is manufactured from polished cobalt-chrome alloy with a constant radius of 24 mm for the tibiofemoral articulation. The component has a trochlear groove with an independent radius for patella tracking. The component is supplied for fixation with either 2 condylar pegs or a short stem, according to surgeon preference. The internal surface is grit-blasted to produce a surface roughness of 3 Ra to which HA coating was added. The tibial component is made of titanium alloy with 2 pegs and an
80-mm stem. The undersurface of the tibial component was HA coated in some cases. The UHMWPE insert was located in the tibial tray. This had the same radius as the femoral component but was sufficiently shallow to allow subluxation at an extreme load. There was no central stabilizer (Fig. 1). Hydroxyapatite Coating Hydroxyapatite coating was applied by plasma spray (Plasma Biotal, Tideswell, UK) to the shotblasted surface of the prosthesis, excluding the condylar pegs or the stem. Its characteristics were purity, ⬎98% HA; crystallinity, ⬎75%; shear strength, 20 to 40 Mpa; and thickness, 80 to 120 m. The surgical technique has been described before [16]. The 2 senior authors (M.A.R.F., G.S.) or trainees under their supervision conducted all the surgeries. Each patient received prophylactic antibiotics, which generally consisted of 3 doses of flucloxacillin. After the first 24 hours, the patients were allowed to bear weight as tolerated using 2 crutches. The crutches were retained for 6 weeks. Follow-Up Evaluations The study cases were followed up prospectively according to a standardized protocol at 6 months and 1, 2, 3, 5, 7, 8, 9,10, and 11 years. At each
846 The Journal of Arthroplasty Vol. 18 No. 7 October 2003
Fig. 2. Nomenclature of different zones of femoral interface.
follow-up evaluation, functional and clinical status were recorded, and weight-bearing anteroposterior and lateral radiographs taken. Twenty-seven of the 36 knees in this series have been regularly reviewed clinically or over the telephone and radiographed as described previously. One further patient (2 knees) remains under review at 11 years but has declined radiography since year 5. Four knees in 2 patients were followed up until the patients died at 7 and 8 years. They are included in the review until then. One patient (1 knee) emigrated at 7 years, and was reviewed until then. This patient was categorized as lost to follow-up at 7 years. These 5 knees were functioning well when last seen. Two patients (2 knees) had radiographs only at 1 year because of severe comorbidity. Both later died at 4 and 8 years. The former knee was not revised at 4 years, but the functional status is not known. The latter knee was functioning satisfactorily on the basis of telephone review at 8 years. These patients were included in the review until their deaths. Radiologic Evaluation The follow-up radiographs were evaluated for alignment of the components to the tibia and femur, overall alignment, and femoral component loosening. The alignment was studied as per criteria described by the American Knee Society [17]. In the lateral view, the femoral interface was divided into different zones (Fig. 2). In the anteroposterior view, lucencies and osteolysis were investigated around the stem when present. The following radiologic signs were taken as signs of loosening. These are based on an earlier study conducted in the same unit on uncemented knee arthroplasties [18]:
i) A radiolucent line more than 2 mm in thickness. ii) Progressive radiolucent lines. The lines were considered progressive if the thickness increased or if it progressively involved the adjacent zones. iii) Proximal migration of femoral component manifest by medial and lateral bone protrusions in the anteroposterior view or increased distance between the posterior lip of the femoral prosthesis and the tip of posterior condyle in the lateral view. iv) More than 3° change in the position of the femoral component in the anteroposterior or lateral view. The results were compared with a previously published report of cemented and cementless (nonHA) fixation of F/S components. These patients also underwent surgery performed by the senior authors using the same surgical technique for the same clinical indications. These procedures differed only in the manner of femoral component fixation [18]. Survival Analysis The life table method was used with revision for aseptic loosening of the femoral component as the criterion for failure to calculate the survival analysis.
Results Thirty-six knees comprised the study group. The demographic data of the group is summarized in Table 1. The average follow-up time was 10 years (standard deviation, 0.83; range, 7–11 years).
Table 1. Demographic Data Age Gender ratio, M:F Side, R:L Diagnosis Primary OA RA Psoriatic arthritis Post-traumatic arthritis Avascular necrosis Previous surgery Patellectomy Arthrodesis Synovectomy Tibial tubercle advancement
Average, 63 y; SD, 9.84 18:18 18:18 30 2 1 1 2 2 1 2 1
Abbreviations: M, male; F, female; R, right side; L, left side; OA, osteoarthritis; RA, rheumatoid arthritis.
Hydroxyapatite-Coated Femoral Components in TKA • Murty et al.
847
Survival Analysis Over the period of the study, 3 revisions have occurred. Two of these were revision of the femoral and the tibial components and one was of the tibial component only. Two cases of aseptic femoral loosening occurred in one patient with bilateral simultaneous arthroplasty. The right knee failed at 1 year and was revised. The other knee failed at 7 years. Both the tibial components were implanted with 5° varus tilt. The femoral components had a reciprocal valgus tilt. At revision of the right side, the femoral component was loose and a complex defect of the medial femoral condyle filled with soft tissue was noted. The tibial component was well fixed with bone growth on the undersurface of the tibial component. which was HA coated. At revision of the left side, the femoral component was again loose and had subsided on the femur, causing erosion of the distal condyles. The UHMWPE insert was partially delaminated. The tibial component in this case was also solidly fixed, and removal was extremely difficult. The third knee developed severe osteolysis of the medial tibial condyle without loosening of the HA-coated tibial component. The UHMWPE insert, which was only 6 mm thick, was found to have severe wear. The femoral component was found stable and was not revised. Therefore, it remains in this review. Histologic analysis showed no evidence of HA particles in the debris, but an abundance of polyethylene particles. Presumably, the osteolysis was caused by accumulated wear debris. The survival rate of HA-coated femoral components at a mean of 7 years after arthroplasty, when 31.5 knees were at risk, was 94% (95% confidence interval, 85.4 –100). At 10 years, the survivorship with revision for any reason is 90.5% (95% confidence interval, 78.5–100). However, the survival of the femoral component, with revision for loosening as the end point, was 94% (95% confidence interval, 84 –100). The survival curve is shown in Fig. 3, and the life table is shown in Table 2.
Fig. 3. The survival curve.
instability. At the final follow-up examination, no objectively demonstrable instability was seen. The patient had previously undergone patellectomy and tibial tubercle advancement. She also had undergone decompression for spinal canal stenosis. No evidence of loosening was seen on radiographs. We cannot explain the instability or the pain. Multiple surgeries causing proprioceptive dysfunction may play a role. The other patient (2 knees) was also symptomatic in the cervical and lumbar spine. Radiographs in this case did not show any evidence of loosening. The pain in this case, too, remains unexplained. One other patient had subjective instability. That patient had undergone no previous surgeries, and the symptoms remain unexplained. Activity Level All our patients were either retired or homemakers. No patient was involved in recreational sports. Uninterrupted Walking Duration Preoperatively, 65% of our patients had a maximum uninterrupted walking duration of more than 10 minutes, of which 7% could walk for more than 30 minutes. At final follow-up evaluation, these figures were 96% and 76%, respectively. The one patient who could not walk had multiple joint involvement because of rheumatoid arthritis.
Clinical Results
Use of Aids for Walking
The clinical results are summarized in Table 3. We did not use a clinical scoring system. We report on the pain level and present the preoperative and postoperative activity level in 7 activities of daily living, namely uninterrupted walking duration, use of aids for walking, ability to get out of a dining chair, ascending and descending stairs, bathing independently and conducting housework. Two patients (3 knees) experienced pain requiring analgesia. One of these patients had associated subjective
In our study, 17% of patients were using either crutches or a frame for walking before the surgery. At final follow-up evaluation, 60% of our patients were walking without the use of any aids, and 36% were using one stick. One patient was using a frame to walk. Ability to Rise from a Chair Preoperatively, 39% of our patients did not report any difficulty in rising from a dining chair,
848 The Journal of Arthroplasty Vol. 18 No. 7 October 2003 Table 2. Survival Analysis (Life Table) Years Since Surgery
Start (n)
Failed (n)
Died (n)
Lost (n)
At Risk (n)
Survival Rate
Confidence Limits
0–1 1–2 2–3 3–4 4–5 5–6 6–7 7–8 8–9 9–10
36 36 35 35 34 34 32 30 27 24
0 1 0 0 0 0 1 0 0 0
0 0 0 1 0 2 0 2 1 0
0 0 0 0 0 0 1 0 0 0
36 36 35 34.5 34 33 31.5 28.5 25.5 20
100 97.2 97.2 97.2 97.2 97.2 94 94 94 94
1.0–1.0 1.0–1.0 .918–1.0 .918–1.0 .918–1.0 .918–1.0 .915–1.0 .856–1.0 .857–1.0 .84–1.0
54% could do it only by pushing off with their hands, and 7% were unable to stand up without help. At final follow-up examination, 63% could perform this activity with ease and 37% could perform this activity with difficulty using a push off.
some help, and 15% either used a shower or were unable to bathe themselves. At final follow-up evaluation, these figures had improved to 60%, 30%, and 10%, respectively. Housework
Ascending and Descending Stairs Before surgery, 86% of our patients could negotiate stairs using a reciprocating gait provided they used a hand rail or climbed stairs with a nonreciprocating gait; 14% were unable to negotiate stairs. At final follow-up evaluation, 84% were climbing and descending stairs using a reciprocating gait .Only 10% needed to use hand rails or had a nonreciprocating gait. One patient was unable to climb stairs, and one patient did not have stairs at home. Bathing Before the knee arthroplasty, 33% of the cohort could bathe without help, 52% could bathe with
Table 3. Clinical Results Pain-free/Occasional analgesia Range of motion
Alignment Instability Stair climbing with reciprocal gait Uninterrupted walking duration
91% FFD Flexion
2 patients (⬍10°) Average, 100° SD, 14° Range, 70°–120° Extensor lag 0 Average 7.2° Valgus SD 1.13° Range 0–10° 2 patients (subjective) 84%
Preoperatively, 11% of our patients were able to do all their housework, 63% were doing part of their housework, and 26% were either unable or had not attempted. At the last follow-up evaluation, the numbers were 63%, 27%, and 10%, respectively. Radiologic Results The mean alignment of the femoral component was 6.9° valgus ⫾ 1.7° (␣ ⫽ 96.9°). The mean alignment of the tibial component was 0.6° varus ⫾ 1.4°. The femoral component was flexed by a mean of 3.8° ⫾ 1.4° (␥ ⫽ 3.8°). The slope of the tibial component was directed posteriorly by a mean of 1° ⫾ 2.1° (␦ ⫽ 89°).The mean tibiofemoral alignment was 7.2° ⫾ 1.1° valgus. For 32 knees, we had a radiograph from at least a 5-year follow up. Of these, 12 knees had no identifiable radiolucent lines. In the remaining 20 knees, 30 radiolucent lines were seen (Fig. 4). None of these were progressive, more than 2 mm in thickness, or associated with other signs of prosthetic loosening. On the basis of the criteria described, no loose femoral components were noted. However, because the radiographs taken were not under fluoroscopic control, some radiolucent lines may have been hidden (Fig. 5).
⬎30 min
76%
Discussion
10–30 min ⬍10 min
20% 4%
This study reports 94% survival of the HA-coated femoral component in total knee arthroplasty at an
Hydroxyapatite-Coated Femoral Components in TKA • Murty et al.
Fig. 4. Distribution of radiolucent lines in different zones.
average follow-up time of 10 years, with femoral revision for aseptic loosening as the end point. Although these results are encouraging, they are not as good as those for cemented knee arthroplasties. We have concentrated this report on the survival of the femoral components only because only 16 patients had HA coating for both components. From this subgroup, 3 tibial components have been revised, but none were found to be loose at revision. Two revisions were performed in 1 patient when the femoral component was found to be loose, but the tibial component was solidly fixed with bone growth on the undersurface of the tibial component. The third revision was for osteolysis of the medial tibial condyle. The tibial component was not loose and was extremely difficult to remove. In this small experimental study, not all component sizes were available. The allocation of the femoral component (whether pegged or stemmed) was determined by availability. Similarly, the correct tibial size was not always available with HA coating. Therefore, some patients received a cemented tibial component. Aseptic loosening of the femoral component is generally believed to be less frequent than that of the tibial component in total knee arthroplasty [19]. The failure rate of uncemented femoral components was been reported to be between 0% and 5% in some short- and long-term follow-up reports [3,20,21]. However, other studies have reported up to 15% failure at 5 years [22]. A previous study on uncemented non–HA-coated F/S femoral components conducted in this unit reported 10% failure rate at 7 years [18]. Thus, the survival of F/S HA-coated femoral components seems to be better than that of most other uncemented prostheses in the medium term. The stability of any prosthesis depends on the interlock achieved between the bone and the im-
849
plant surface [23]. Researchers have shown that the cuts made by a saw are not uniformly smooth [24], and gaps are inevitable between the implant and the bone surface. Cementing under pressure fills these gaps and provides the necessary interlock, which can resist shear and rotary stresses [23]. HA can induce chemical bonding, initially providing the necessary interlock [10]. HA can also conduct bone growth onto the surface of the implant across a gap up to 5 mm wide [6,7]. It has been shown to enhance stability by converting fibrous tissue to bone under load [25]. Animal experiments have shown that in a cancellous site HA-conducted bone growth to the implant occurs in 4 to 6 weeks [26]. Provision of initial stability to enable the osteoconductive properties of HA to take effect seems to be important; micromotion of 150 m has been shown to inhibit bone formation [27]. In our series, one of the 2 knees that failed did so early, lending credence to this hypothesis. The other knee in the same patient failed after 7 years .The alignment of the prostheses in these 2 knees was poor, producing an oblique joint line and presumably abnormal interface loading. This may have lead to failure, whatever the method of fixation. However, even in these knees, both components showed evidence of bone ongrowth (Fig. 6). A reported complication of HA coating at the hip is the generation of particulate matter leading to third body wear. HA particles found in osteolytic areas are associated with an inflammatory reaction [28]. One of our knees had severe osteolysis in the tibial condyle, which required tibial revision. The tibial and femoral components were well fixed. The UHMWPE insert, which was only 6 mm in
Fig. 5. Anteroposterior and lateral radiographs of a typical patient at 6 years after surgery.
850 The Journal of Arthroplasty Vol. 18 No. 7 October 2003
Fig. 6. (A) An ultraviolet photograph of the internal surface of the femoral component. The U-V light has fluoresced the bone and appears bright white. The light grey areas are the HA coating without bone ongrowth and the dark grey area show where the HA coating is absent. (B) An ultraviolet photograph of the undersurface of the tibia tray of the component. As in panel A, bright white areas are ongrown bone; light grey areas are exposed HA without ongrowth, and dark gray areas are where the HA is absent.
initial thickness, showed severe wear. On microscopy of the contents of the osteolytic cavity, polythene debris was abundant but HA particles were not detected. Unfortunately, the polythene component was not further examined, so that we have no definite explanation for severe wear in this case. Because the femoral component was well fixed and was not revised, this knee did not represent a failure as defined here. HA did not seem to be directly responsible for osteolysis, and HA fixation withstood a large polythene load. Radiostereophotogrammetric analysis (RSA) has shown that migration of HA-coated tibial components is comparable to that with cement over 2 to 5 years [14,29]. The stability with regard to micromotion and migration is superior to that of non–HAcoated implants [30]. In light of these findings, we believe that HA-coated TKA should be evaluated with guarded optimism because it has the potential for conserving bone stock. However, in our experience to date, cemented fixation continues to be the most reliable form of fixation.
2.
3.
4.
5.
6.
7.
8.
References
9.
1. Robertsson O, Scott G, Freeman MAR: Ten year survival of the cemented Freeman Samuelson pri-
10.
mary knee arthroplasty. J Bone Joint Surg Br 82:506, 2000 Scuderi GR, Insall JN, Windsor RE, Moran MC: Survivorship of cemented knee replacements. J Bone Joint Surg Br 71:798, 1989 Fanning JW, Joseph J Jr., Kaufman EE: Follow up on uncemented total knee arthroplasty. Orthopaedics 19:933, 1996 Hungerford DS, Krackow KA, Kenna KA: Cementless total knee replacement in patients 50 years old and under. Orthop Clin North Am 20:131, 1989 Petty W: The effect of methylmethacrylate on bacterial phagocytosis and killing of human polymorphonuclear leucocytes. J Bone Joint Surg Am 60:752, 1978 Soballe K, Hansen ES, Brockstedt-Rasmussen H, et al: Gap healing enhanced by hydroxyapatite coating in dogs. Clin Orthop 272:300, 1991 Stephenson PK, Freeman MAR, Revell PA, et al: The effect of hydroxyapatite coating upon the in growth of bone into cavities in an implant. J Arthroplasty 6:51, 1991 Bauer TW, Geesink RCT, Zimmerman R, McMahon JT: Hydroxyapatite coated femoral stems. J Bone Joint Surg Am 73:1439, 1991 Furlong RJ, Osborn JF: Fixation of hip prosthesis by hydroxyapatite ceramic coatings. J Bone Joint Surg Br 73:741, 1991 Geesink RGT, De Groot K, Klein CPAT: Bonding of
Hydroxyapatite-Coated Femoral Components in TKA • Murty et al.
11.
12. 13.
14.
15.
16.
17.
18.
19.
20.
bone to apatite coated implants. J Bone Joint Surg Br 70:17, 1998 Geesink RGT, Hoefnagels NHM: Six- year results of hydroxyapatite coated total hip replacement. J Bone Joint Surg Br 77:534, 1995 Kroon PO, Freeman MAR: Hydroxyapatite coating of hip prostheses. J Bone Joint Surg Br 74:518, 1992 McNally SA, Shepperd JAN, Mann CV, Walczak JP: The results at nine to twelve years of the use of a hydroxyapatite coated femoral stem. J Bone Joint Surg Br 82:378, 2000 Nillson KG, Karrholm J, Carlsson L, Dalen T: Hydroxyapatite coating versus cemented fixation of the tibial component in total knee arthroplasty. J Arthroplasty 14:9, 1999 Onsten I, Nordquist A, Carlsson AS, et al: Hydroxyapatite augmentation of the porous coating improves fixation of tibial components. J Bone Joint Surg Br 80:417, 1998 Freeman MAR, Samuelson KM, Bertin KC: Freeman-Samuelson total knee arthroplasty. Clin Orthop 192:46, 1985 Ewald FC: The knee society total knee arthroplasty roentgenographic evaluation and scoring system. Clin Orthop 248:9, 1989 Chokalingham S, Scott G: The outcome of cemented versus cementless fixation of a femoral component in TKR with the identification of radiological signs for the prediction of failure. Knee 7:233, 2000 Insall JN, Haas SBComplications of total knee arthroplasty. p. 891Surgery of the knee2 EdIn , Insall JN, Windsor RE, Scott WN, et al. New York, Churchill Livingstone, 1993 Moran CG, Pinder IM, Lees TA, Midwinter MJ: Sur-
21. 22.
23. 24.
25.
26.
27.
28.
29.
30.
851
vivorship analysis of uncemented porous-coated anatomic knee replacement. J Bone Joint Surg Br 73: 848, 1991 Whiteside L: Cementless total knee replacement. Clin Orthop 309:185, 1994 Campbell MD, Duffy GP, Trousdale RT: Femoral component failure in hybrid total knee arthroplasty. Clin Orthop 356:58, 1998 Freeman MAR: Hydroxyapatite catalyst or conjurer? Orthopaedics 21:959, 1998 Toksvig-Larsen S, Ryd L: Surface characteristics following tibial preparation during total knee arthroplasty. J Arthroplasty 9:63, 1994 Soballe K, Hansen ES, Brockstedt-Rasmussen H, Bunger C: Hydroxyapatite coating converts fibrous tissue to bone around loaded implants. J Bone Joint Surg Br 75:270, 1993 Stevenson S, Goldberg VM, Tisdel CL, Dean JC: Application of hydroxyapatite in total knee arthroplasty. p. 99. In Insall JN, (ed). Current concepts in primary and revision total knee arthroplasty New York, Lipincott-Raven, 1996 Soballe K: Hydroxyapatite ceramic coating for bone implant fixation: mechanical and histological studies in dogs. Acta Orthop Scand Suppl 255:1, 1993 Bloebaum RD, Beeks D, Dorr LD, et al: Complications with hydroxyapatite particulate separation in total hip arthroplasty. Clin Orthop 298:19, 1994 Toksvig-Larsen S, Ryd L, Jorn L P, et al: Hydroxyapatite enhanced tibial prosthetic fixation. Clin Orthop 370:192, 2000 Nelissen RGHH, Valstar ER, Rozing P: Effect of hydroxyapatite on micromotion of total knee prostheses. J Bone Joint Surg Am 80:1665, 1998