Examination of surface and material properties of explanted zirconia femoral heads

The Journal of Arthroplasty Vol. 19 No. 7 Suppl. 2 2004

Examination of Surface and Material Properties of Explanted Zirconia Femoral Heads Erick M. Santos, MD, PhD,* Shikhar Vohra,† S. Aaron Catledge, PhD,† Michelle D. McClenny, BSN,* Jack Lemons, PhD,* and K. David Moore, MD*

Abstract: The crystalline structure of Zirconia femoral heads provides for superior fracture toughness when compared with alumina. Transformation of the crystalline structure that takes place with time in service may produce changes in the surface and biomechanical properties of these implants. This study examines surface and mechanical property changes of Zirconia femoral heads that occur with time in situ. Eighteen retrieved Zirconia femoral heads were compared to 5 factory-sealed controls. The retrieved implants demonstrated significant transformation to a monoclinic phase. This phase transformation was associated with decreased surface hardness. There was evidence of increased surface roughness with increasing time of implantation. The phase transformation that takes place in Zirconia femoral heads may render these implants less desirable as a bearing surface in total hip arthroplasty. Key words: zirconia, ceramic, hip, prosthesis, implant, wear. © 2004 Elsevier Inc. All rights reserved.

The use of alternative bearing surfaces in total hip arthroplasty (THA) has been advocated as a way to prolong the life of implants by decreasing wear between components [1,2]. Ceramics, such as Tetragonal Zirconia Polycrystal (TZP) and Alumina have been used as materials for femoral heads in THA because of their desirable mechanical properties (high fracture toughness, surface hardness, and low coefficient of friction) [3]. TZP femoral heads were introduced in 1985 as an alternative to alumina, because of zirconia’s superior hardness and resistance to fracture [4]. However, there have been recent reports of significant polyethylene wear with TZP ceramic femoral heads [5,6].

TZP ceramic exists in three different phases of crystalline structure (monoclinic, tetragonal, and cubic) [7]. Figure 1 describes the three dimensional shape of the arrangement of atoms between zirconium and oxygen in the lattice structure that composes zirconia at different temperatures and different contents of stabilizing atoms (such as yttrium). TZP as it is manufactured is in a metastable tetragonal phase that has the advantage of possessing higher resistance to crack formation than the monoclinic phase, which is the stable phase at room temperature [7]. TZP in a tetragonal phase can undergo transformation into the monoclinic phase under stress or with aging [4,7–9]. The phase transformation from tetragonal to monoclinic involves a volume expansion of 3– 4%; because of this volume expansion, crack propagation is halted in the tetragonal phase as the crack is sealed [4,7]. Once the material undergoes extensive transformation into a monoclinic phase, this advantage is lost and the implant may be more susceptible to surface damage and to increasing surface roughness [9]. It has been postulated that in an in vivo environment, TZP ceramic femoral heads undergo aging

From the *Division of Orthopaedic Surgery and the †Department of Physics, University of Alabama-Birmingham, Birmingham, Alabama. No benefits or funds were received in support of the study. Reprint requests: K. David Moore, MD, 920 Faculty Office Tower, 510 20th Street South, Birmingham, AL 35294. © 2004 Elsevier Inc. All rights reserved. 0883-5403/04/1907-2006$30.00/0 doi:10.1016/j.arth.2004.06.017

30

Surface Changes of Zirconia Femoral Heads • Santos et al.

Fig. 1. Diagram of the crystalline structure of TZP in its tetragonal and monoclinic phases.

with increasing conversion of their surface layers into monoclinic phase [4,8,9]. Over time, this phase transformation could lead to increasing crack formation and decreased surface hardness, which in turn would leave the surface vulnerable to wear from increasing surface roughness. There have been two recent studies that have examined a limited number (2 in one case, 3 in the other) of explanted TZP femoral head samples for signs of increased wear, but no study has examined a large number of samples for signs of surface deterioration [6,9]. The purpose of our study was to examine the surface and bulk mechanical properties of explanted TZP femoral heads compared to factorysealed controls. The hypothesis is that degradation of surface and mechanical properties occurs with increasing time of implantation because of the surface material phase transformation from tetragonal to monoclinic phase. The aim is to compare the control specimens to explanted samples that have been subjected to the rigors of implantation and determine the extent of degradation, if any, which occurs over time in the human body.

31

Japan: 6, as confirmed via their laser etch markings present on the Morse taper. Eighteen TZP femoral heads and 5 factory-sealed controls (Ceraver, France) were examined using light microscopy (LM), scanning electron microscopy (SEM), glancing angle x-ray diffraction (XRD), laser profilometry (LP), and nanoindention (NI) hardness testing. Samples were subjected to ultimate compression loading (UCL) in a hydraulic press. All samples were examined under light microscopy at magnifications of ⫻10 and ⫻100 to identify any areas of gross wear, and SEM (ISI-30B model manufactured by International Scientific Instruments, Santa Clara, CA) was performed to evaluate areas of wear and surface topography. Surface hardness was evaluated with the Nanoindenter instrument (Nano Indenter XP model, MTS Systems, Oak Ridge, TN). Tetragonal and monoclinic phase composition of the samples was examined using glancing angle XRD (X’Pert MPD, Philips, Eindhoven, The Netherlands). Peaks from the XRD output were compared to library controls and a ratio of the peak strength (area under the peak) was used to calculate tetragonal to monoclinic ratios. The units are in millimol % and are derived by taking the ratio of tetragonal to monoclinic peaks areas in XRD and comparing that to a known library control of either 100% monoclinic or 100% tetragonal content in millimol percent units. A laser profilometer (Alpha Step 500 Surface Profilometer, Tencor Instruments, Milpitas, CA) was used for surface roughness measurements (Ra). UCL testing was done on all samples by mounting them on their original femoral stems and placing them in an industrial hydraulic press until they failed catastrophically.

Results Materials and Methods Samples for our study were collected after revision total hip arthroplasty in our implant retrieval center, separated from their femoral stems and washed in an acetone bath to remove organic residue. Complete clinical data was available for 5 of the 18 explants. Partial data including height, weight, age, sex, and time of implantation was available for an additional 6 specimens. For the samples in which clinical data were available, the mean time of implantation 63.6 months (range, 2–120). The samples had the following manufacturers: Ceraver, France: 7; Astromet, USA: 5; Kyocera,

LM and SEM revealed evidence of metallic transfer, increased surface roughness and pitting in explanted samples. These changes were more pronounced at the weight-bearing surfaces near the pole to 45° above the equator. Figure 2A shows a control surface with almost no surface defects seen at ⫻1,000 magnification with SEM. Figure 2B shows a sample implanted for 62 months with a typical area of wear on the weight-bearing surface under the same magnification with SEM. XRD and NI testing showed good correlation between increasing surface monoclinic phase and decreasing surface hardness. Control samples had a mean monoclinic mmol percent content of 3.4%,

32 The Journal of Arthroplasty Vol. 19 No. 7 Suppl. 2 October 2004

Fig. 2. (A) Scanning electron microscopy at 100x showing almost no surface defects in a control specimen. (B) Weight-bearing surfaces of a sample implanted for 62 months at the same magnification.

while the mean for explanted samples was 21.49%. Samples with longer times of implantation had increasing monoclinic phase conversions as seen in Figure 3. Laser profilometry data showed a rough correlation (r2 ⫽ 0.04) between surface roughness and percent monoclinic content as seen in Figure 4. The best correlation (r2 ⫽ 0.44) was shown for in-

creased surface roughness and time of implantation as seen in Figure 5. The trend was for increasing surface roughness with time. UCL strength averaged 561.25 MPa with standard deviation (SD) of 344.14 for explanted samples; controls averaged 1,253.4 MPa with SD of 850.18. While no correlation was found between time of implantation and compression strength, in

Fig. 3. X-ray diffraction results versus time of implantation that demonstrate increasing percentage of monoclinic phase with increasing time in situ.

Fig. 4. Correlation of surface roughness with percentage of monoclinic phase.

Surface Changes of Zirconia Femoral Heads • Santos et al.

Fig. 5. Surface roughness compared with time of implantation.

general, there was a trend for the control samples to have higher strength than the explanted samples.

Discussion TZP implants show varying degrees of monoclinic phase transformation with time of implantation. The surface phase correlates with hardness and can be fit in a quadratic equation, as shown in a previous study by our research group [10]. Surface phase transformation shows no correlation with UCL strength, and only a rough correlation with surface roughness. The difference between UCL results between control samples and explanted samples can be explained because of microcrack formation on the surface of the explanted samples. These cracks and other surface imperfections are demonstrated with our SEM survey. With increasing time of implantation, there is a clear degradation of the surface with increased surface roughness, as shown by our SEM and laser profilometry examinations. While there are many variables involved in the degradation of TZP implants in vivo, there is evidence that the surface phase transformation plays an important role. Our study confirms the role of the surface phase transformation, but from our results it can be seen that this is not the only factor

33

affecting surface degradation in vivo. There was not a clear correlation between increasing monoclinic surface content and surface roughness, and this seems to suggest that other factors, such as third body wear or contact with the acetabular shell during reduction maneuvers most likely play crucial roles in the long-term health of the TZP surface. There have been previous studies that have shown surface transformation in TZP under the influence of cycling and pressure under simulated physiological conditions [7]. There have also been case reports of failures of TZP implants in vivo with impingement of the TZP ball on metal acetabular liners [4]. All of these effects can lead to the surface defects we have detected in this study with even modest increases in temperature and pressure over the course of years producing enough surface wear to lead to problems such as osteolysis. A recent study compared TZP femoral head implants with stainless steel and alumina, finding that there is an increase in wear over time for TZP samples and examining a limited number of retrieved samples for surface degradation [6]. Their results, while limited to 3 explanted samples, agree with our examinations. A weakness of this study is the lack of complete clinical data patients from whom the femoral heads were retrieved. Maximum efforts were undertaken to retrieve this clinical data. However, our lab received the specimens from a wide variety of sources across our geographic region over a number of years. Unfortunately no clinical data were available for 7 implants and only limited data were available for 6. Factors such as patient weight, age, sex, activity level, history of dislocations, and malpositioned components may certainly have contributed to surface changes on the explanted femoral heads. In addition, not all samples were able to be tested in the same sequence; most samples were examined first with LM, and then subjected to UCL testing, then SEM, XRD, NI, and LP. This may have affected our results, but the area of distortion from UCL was small and confined to an area of about 5 mm in diameter at the northern pole of the samples. Prior to UCL testing, the samples had been marked to identify the areas of further examination, and the UCL testing yielded for the most part large pieces that did not have any visible changes or cracking under LM after the test. Variations between manufactures could account for some of the differences between controls and retrieved specimens, particularly in terms of compressive strength; however, the non-weight-bearing surface of each implant also served as an internal control in terms of changes in surface roughness.

34 The Journal of Arthroplasty Vol. 19 No. 7 Suppl. 2 October 2004 While TZP ceramics possess superior hardness and compression strength than alumina on implantation, the phase transformation to a less desirable monoclinic phase in TZP over time makes this material less suitable for the demands of THA than materials such as alumina that do not undergo this type of phase transformation.

Acknowledgment The authors wish to acknowledge the invaluable assistance of Yogesh Vohra, PhD and Monique Cook, MS in carrying out this research.

4.

5.

6.

7. 8.

References 1. Inzerillo VC, Garino JP: Alternative bearing surfaces in total hip arthroplasty. J South Orthop Assoc 12(2): 106, 2003 2. Jazrawi LM, Kummer FJ, DiCesare PE: Alternative bearing surfaces for total joint arthroplasty. J Am Acad Orthop Surg 6(4):198, 1998 3. Christel P, Meunier A, Heller M, et al: Mechanical properties and short-term in-vivo evaluation of yt-

9.

10.

trium-oxide-partially-stabilized zirconia. J Biomed Mater Res 23(1):45, 1989 Cales B: Zirconia as a sliding material: histologic, laboratory, and clinical data. Clin Orthop 379:94, 2000 Simon JA, Dayan AJ, Ergas E, et al: Catastrophic failure of the acetabular component in a ceramicpolyethylene bearing total hip arthroplasty. J Arthroplasty 13(1):108, 1998 Hernigou P, Bahrami T: Zirconia and alumina ceramics in comparison with stainless-steel heads: polyethylene wear after a minimum ten-year follow-up. J Bone Joint Surg Br 85(4):504, 2003 Piconi C, Maccauro G: Zirconia as a ceramic biomaterial. Biomaterials 20(1):1, 1999 Cales B, Stefani Y, Lilley E: Long-term in vivo and in vitro aging of a zirconia ceramic used in orthopaedy. J Biomed Mater Res 28(5):619, 1994 Haraguchi K, Sugano N, Nishii T, Miki H, Oka K, Yoshikawa H: Phase transformation of a zirconia ceramic head after total hip arthroplasty. J Bone Joint Surg Br 83(7):996, 2001 Catledge SA, Cook M, Vohra YK, et al: Surface crystalline phases and nanoindentation hardness of explanted zirconia femoral heads. J Mater Sci Mater Med 14:863, 2003