Zirconia ceramics
or "by night, all cats are grey"

A Dambreville*, M Phillipe**, A Ray***
*Clinique St-Michel, 29000 Quimper
** Hôpital de Carpentras 84200 CARPENTRAS
***Clinque du Parc 69458 LYON CEDEX - France

In the absence of rigorous scientific clarification, information on bio-materials is frequently a source of confusion and misleading generalisations worrying to orthopaedic surgeons. A recent exemple : total hip replacements in titanium. Some surgeons observed osteolyses around the stems of femoral prostheses in cemented titanium and concluded his combination should no longer be used, forgetting many other surgeons have successfully been sealing titanium prostheses for more than twenty years. All that was needed was to realise some surgeons were using stems in titanium with a rough surface and other stems in an anodised smooth titanium for the confusion to be cleared up (7). But as in darkness all cats appear grey so under a blanket of ignorance all titanium prostheses seemed the same.

Another recent example is hydroxyapatite (HAP). In 1991 Blocbaum, a well-known American surgeon, published an article on premature osteolyses with prostheses coated with hydroxyapatite (HAP). This caused problems in the USA where use of HAP was no longer recommended until a study in Europe concerning thousands of cases over a ten year period proved there is HAP and HAP and if performed correctly, this surface treatment is highly effective (4). Today HAP is approved by the FDA. Although it is useful and even essential for surgeons to report on their failures, the information must be comprehensive, allowing the reader to make a critical assessment in the knowledge of the precise properties of the material used.

The topic of concern to us in this article is zirconia ceramics.

Zirconia has been used in orthopaedics since 1985. The manufacturing processes have evolved and are extremely accurate. Depending on the quality of manufacturing, the properties of the zirconia balls obviously vary - as do the results observed : there is zirconia and zirconia.

10 years ago several companies were seeking to control the technology . In view of the unreliability of the products manufactured most gave up. The main companies currently manufacturing zirconia heads are five world-wide : two in Japan (Kyocera and NGK), three in Europe : Morgan-Matroc in England, Metoxit in Switzerland and Norton - Desmarquest (Prozyr heads) in France. Only the last three companies sells zirconia heads in France. What is zirconia ? What criteria make a difference ? We will now attempt to answer these questions.

WHAT IS ZIRCONIA ?

A few definitions :

Zirconium : white metal (Zr), atomic number 40, density 6.51.

Zirconia : zirconium oxide (ZrO2).

Ceramic : poly-crystalline material produced by sintering.

Sintering : densification of powders by heating.

Yttrium : metal (Y), atomic number 39, included in the rare earth group.

A zirconia head is constituted of zirconium oxide ceramic.
Its quality depends on its purity, density, porosity, particle size, crystalline structure (content of tetragonal and monoclinic phases), bending strength, geometric characteristics, tolerances and surface condition.

Purity (Figure 1) :
Zirconium oxide powder is purified by a chemical reaction of dissolving-precipitation.To obtain satisfactory stability, the presence of yttrium oxide is necessary, the ideal proportion being 5.15 wt.%. The most common impurities are aluminium oxide or alumina , the proportion of which should be less than 0.5 wt.%.

Figure 1

Density :
The density of the zirconia should be as close as possible to the theoretical density (100% dense), which is 6.1. The closer it is to this value, the fewer the spaces between the particles and the greater the mechanical strength and the less the surface roughness. The norm is 6 and the ideal 6.1.

Porosity :
To use zirconia under friction, in particular, for prosthesis heads, the porosity must be as close as possible to zero to obtain the lowest possible surface roughness. Any manufacturing defects, in particular, use of large particle powders will promote wear by abrasion. Particle size :
The ceramic particles should be smaller than 0.6mm. The strength of the balls and their surface condition are governed by the particle size.

Tetragonal and monoclinic phases :
The zirconium oxide crystals are arranged in crystalline cells (mesh) (Figure 2). These may be cubic with square sides, the cubic phase ; in the form of a straight prism with rectangular sides, the tetragonal phase, or in the form of a deformed prism with parallelepiped sides, the monoclinic phase (Figure 2). The cubic phase, formed at high temperature, has moderate mechanical properties. Only the tetragonal phase allows obtaining a ceramic with satisfactory mechanical properties. The monoclinic phase weakens the mechanical performance and may contribute to a reduction in the cohesion of the ceramic particles and thus of the density. Thus it is clearly essential for the monoclinic phase to be as small as possible (Figure 1). We will see the manufacturing and sterilisation processes should restrict the potential for transformation from a tetragonal to a monoclinic phase.

Figure 2

Bending strength :
The standards (Figure 1) for mechanical strength of the zirconia when subject to a bending force applied at four points, is at least 800 Mpa (mega pascals) (10Mpa = 1 kg/mm2). The mechanical strength depends on density, particle size, the content of tetragonal phase, .... A satisfactory fatigue strength of good quality zirconia heads has been proven.

Geometry :
Obviously, the spherical shape must be as near to perfect as possible (tolerance less than 10 mm). The stated dimensions must be extremely accurate (tolerance for a diameter of 28 : ±10mm). The geometric characteristics are governed by the final production stages.

Surface roughness :
The wear behaviour depends on the roughness of the spherical part of the head. The surface roughness is generally expressed as Ra (mean distance between the valleys and peaks in relation to mid-line) (Figure 3). The standard imposes an Ra of less than 0.03mm. This standard now appears inadequate. Although the heads supplied some five years ago satisfied this norm, those currently used have much improved performance standards with an Ra between 0.002 and 0.003 mm.

Figure 3

We have just reviewed the criteria which characterise zirconia heads. Each factor is important . As for any other component, it is not sufficient to write the word "zirconia" to inform the reader of the exact nature of the product. All the criteria must be indicated. They depend on the quality of manufacturing. To understand the differences, it is useful to understand the various manufacturing phases.

MANUFACTURE OF ZIRCONIA HEADS :

The zirconia powder is compacted and subject to an initial heat treatment (sintering). It is then subjected to a second treatment, HIP (hot isostatic pressing) followed by treatment at a lower temperature (whitening). The ceramic produced is then machined and polished.

The zirconium oxide powder must be as pure as possible and the particle size sufficiently small (crystals of 0.1mm). It is evenly mixed with yttrium oxide to allow stabilizing of the zirconia crystals and promote the tetragonal phase (Figure 2). The proportion of yttrium oxide must be very accurate : 5.15 wt.% and its distribution totally uniform.

1 Compacting :
The particles are compressed under extremely high pressure, in the order of 1000 bar, to initiate adhesion. The manufacturing secrets, intensity of the pressure, and conditions of application govern the quality of production.

2 Sintering :
Sintering, which is the binding together of the particles, occurs at a temperature of less than 1500°C. The temperature for industrial sintering is between 1400°C and 1500°C. Over 1500°C there is a risk of formation of the cubic phase which does not possess satisfactory mechanical characteristics. Below 1400°C the risk of poor density arises. Between 1400°C and 1500°C, the lower the temperature, the smaller the particles, which is what is required. Each manufacturer has its manufacturing secrets for obtaining the best possible ceramic.

During sintering, the zirconia part shrinks by 20% to 30% which dictates starting out with over-sized blanks. For example, a ball of diameter 40 mm is required to produce a 28 mm head after sintering.

3 HIP (Hot Isostatic Pressing) :
To improve the density of the particles, still further, the zirconia ceramic is heated to between 1400°C and 1500°C under high pressure (>=1000 bar) in an inert atmosphere. This treatment produces an ideal density of 6.1 g/cm3.

Before 1995, not all heads were subject to HIPing, hence their quality was not that achieved today.

4 Whitening :
Since HIPing is performed under argon and in the absence of oxygen, the balls acquire a dark grey coloring. Whitening is achieved by heating in air. Once again, manufacturing secrets govern this stage.

5 The first controls :
The density is checked at this stage.

6 Grinding or "rounding off" :
The balls are placed in grooved trays or run around with diamond paste. This process produces almost perfect spheres.

7 Polishing :
By using the same process as for grinding but with a paste containing finer diamond particles, the balls are polished. This treatment governs the surface roughness.

8 The second controls :
The geometric and surface characteristics of the balls are checked.

9 Machining :
Only the cones are made by machining under high lubrication. Of course, the adjusting of the taper of the stem to the bore of the head must be performed with extreme accuracy.

STERILISATION OF THE HEADS

Some years ago rapid wearing of polyethylene with zirconia heads was reported . The most likely cause of these incidents was sterilization in an autoclave. Such sterilisation may generate an excess percentage of the monoclinic phase (Figure 2). At the time, this gave rise to official memos, in particular, in France and the USA. This problem concerned old heads which were not subjected to HIPing.

The quality of heads undergoing 100% HIPing does not give rise to the same disadvantages. However, zirconia heads must be sterilized by beta or gamma radiation or ethylene oxide and never in an autoclave.

RADIOACTIVITY OF ZIRCONIA HEADS

Some zirconium oxide powders contain radioactive impurities which during the early 1990’s gave rise to fears of substantial radioactivity of zirconia heads. Manufacturers currently use very pure powders and the radioactivity of zirconia heads is very low (100gy/h) as comparable to that of alumina or metal heads (8).

AGEING OF ZIRCONIA

Zirconia has been questioned on instability over time, with a transformation from the tetragonal phase to the monoclinic form (Figure 2). Again, this depends on the quality of the zirconia. The heads of the quality now achieved, taken as references and which have benefited from all the manufacturing processes described above, are regularly tested in an autoclave for 5 hours at 134°C to simulate 20 years ageing in vivo. This treatment generates only 2 to 3% of the monoclinic phase, which is low and apparently without any disadvantages. It has been established in vitro (2) that the presence of less than 40% monoclinic phase does not result in any measurable increase in polyethylene wearing.

LABORATORY TESTS

The strength tests of zirconia have confirmed its high breaking strength.

Wear tests against polyethylene cups have confirmed low wear rates comparable to those for alumina and substantially lower than those of metals (6).

CLINICAL ASPECTS

Clinical studies achieved with zirconia heads are rare.

How can a valid clinical study be conducted ? Detection of indirect signs of polyethylene wear, femoral or cotyloid osteolysis incorporate bias, since they depend on several factors : design and composition of the femoral or cotyloid prosthesis, quality of the polyethylene, of the cement, age of the patients, their activities. A comparative study of a series of comparable patients in terms of their age, sex, aetiology, ... with identical prostheses where only the composition of the head was different, with heads of the same diameter and an identical time lapse would be ideal. To our knowledge no such study has never been performed.

Another difficulty is that a sufficient time lapse is necessary to obtain significant results. Moreover, the quality of zirconia has changed radically since the beginning and a study of old heads would not be a true image of the performance achieved by the heads currently being manufactured.

Doctor Le Mouël, a member of the Mondor hospital team (Prof. Goutallier) was a pioneer in this field and observed a series of very worrying osteolysis problems with an old series of zirconia heads. He reported during a SOFCOT meeting (9) and agreed to participate in this work. He describes his experience : "The zirconia heads used between December 1988 and February 1991 were from the company CERAMIQUE ET COMPOSITE (RHONE POULENC) which has currently ceased business. The zirconia used was HIP treated since the beginning. This sintering under argon, used by all manufacturers since it improves the mechanical performances of the zirconia, results in a loss of O2 molecules from the surface of the zirconia after sintering, giving it a black color. Now zirconia manufacturers re-oxidise the zirconia to obtain a white or violet coloring. It was black HIP treated zirconia that we were using at that time.

We encountered problems of femoral or cotyloid osteolyses when reviewing the 100 first total hip replacements using prostheses with a combination of Ø 28 zirconia heads and polyethylene. This was a consecutive series made between January 1988 and January 1991. 78 prostheses reviewed after more than one year were taken into account. These had been inserted in 61 patients (17 bilateral total hip replacements) of average age 66. The femoral implant was a smooth cemented titanium (Céraver), the cement was either Palacosgenta or Cérafix (opacifying agent: zirconium dioxide). The cup was cemented polyethylene (Céraver). We found 11 prostheses (14%) presenting aseptic loosening, of which 2 femoral (2.5%) and 10 of the cup (13%) (including one case of bilateral loosening). Loosening occurred from the 6th year. We also noted abnormal osteolysis from the 4th year in the diaphysic femoral implant in 20% of the femoral implants in Gruen and Amstutz zones 2-3-5-6. We also found a calcar osteolysis (zone 7) in 51.3% of femurs with an average height of 5.5 mm (minimum 2.5, maximum 12). The same titanium femoral implant from Céraver was used in the hospital since 1979 and was associated with metal-polyethylene and alumina-polyethylene combinations (head Ø32), with cemented cups. We investigated these first two series with an identical population, and obtained the following results (10) :

metal-polyethylene series : 16% loosening of the cup and 1.3% loosening of the femoral implant.

alumina-polyethylene series : 2.6% loosening of the cup and 0.85% femoral loosening, 30% of femurs presented osteolysis at the calcar and 4.25% diaphysic osteolyses.

The satisfactory results for the smooth titanium implant were confirmed in these two combinations and confirmed by the studies of St Louis hospital (Nizard) and Lariboisière hospital as well as of the hospital in Rennes (Fr). Thus it was totally anomalous to find these images in the zirconia-polyethylene series. There was no abnormal wear of polyethylene. The measurements made on our series resulted in average linear wear of less than 0.1 mm per year which appeared less for the head Ø 32 of the alumina-polyethylene combination. We think in this series there is probably a problem of micro-particle bio-tolerance. We are currently conducting bio-tolerance tests on various types of particles, including stabilizing oxides, in cell cultures".

What had happened in the series ? As is frequently the case, the clinical application is more complex than the laboratory study. Did the zirconia originate from the cement or the heads (11) ? Did the absence of whitening promote release of poorly tolerated micro-particles ? Was the zirconia powder used at the time as pure as the powders currently in use ? Was the yttrium oxide critical for stability homogeneously distributed ? Slight variations during the various manufacturing processes, compacting, sintering, HIP could have given rise to the defects. A difference of a few degrees during heating could reduce cohesion of the particles. It is not easy to answer these questions since the precise temperatures and the application conditions are jealously-guarded manufacturing secrets.

Surgeons in the group Futura 2000, and in particular, André Ray and Michel Philippe have encountered totally different experiences with zirconia heads. They studied 101 zirconia head prostheses after a minimum time lapse of 5 years and a maximum of seven years. In all cases, the femoral prosthesis was an Esop prosthesis coated with hydroxyapatite and the cotyloid prosthesis the Esop hydroxyapatite cotyl. The heads were of diameter 28 mm, manufactured by Norton Desmarquest.

In the series of 101 prostheses there were only two cases of cotyloid osteolysis linked to instability of the polyethylene insert in the cup. André Ray measured the polyethylene wear by direct comparative reading of ascension of the centre of the head on the basis of a post-operative X ray and another X ray taken at a later date. 54 measurements were made after 5 years, 31 after 6 years and 16 after 7 years (Figure 5).

Figure 5

Wear was zero in 81 cases, up to 0.5 mm in 10 cases, up to 1 mm in 8 cases and up to 2 mm in two cases (Figure 6).

Figure 6

The last two cases were caused by the loosening of the cup on account of a defect in the insert. Despite these two special cases, average annual wear in this series was clearly less that the normal permissible average of 0.1 mm per year, since it was 0.03 mm per year. These measurements, made using a graduated strip, were relatively inaccurate and slight variations between one measurement and another were possible.

Michel Philippe made more accurate measurements using the Charnley method (1, 3, 5) (Figure 4) on 27 zirconia head prostheses after a time lapse of 5 to 7 years and did not detect any osteolysis.

Figure 4

Average wear was 0.09 mm per year. At the 1997 SOFCOT meeting, during the Symposium on hip replacement before the age of 50, Jenny reported lower wear rates for zirconia heads (0.07 mm per year) than for other components. It may be argued the measurements made on the basis of frontal hip X rays take into account solely ascension of the head in the frontal plane (3). However, all the measurements were convergent and allowed confirming the advantages of zirconia in preventing wearing of the polyethylene.

CONCLUSION

The experience of Dr. Le Mouël confirms the risk of serious problems when introducing new materials and imposes extreme caution. But this should not paralyse our initiatives intended to improve implants. Our clinical experience leads us to believe that after the period necessary for final development, the imporvement achieved allows zirconia to be now considered as a reliable material, provided the qualities of purity, density, porosity, particle size, content of tetragonal phase, bending strength, geometry and surface conditions are satisfactory. The several steps in the manufacturing process must be carefully and accurately controled and followed. The zirconia heads currently approved in France comply with the standards (Table 1) which provides a guarantee. It should be recalled that by law, surgeons must be familiar with the implants they use and may be held liable. To select, in the light of all available knowledge, the correct zirconia head to use, the surgeon is able to question the supplier as to its precise characteristics. This article has perhaps shed some light on the matter.

Bibliography 1 - Charnley J., Halley DK (1975) Rate of wear in total hip replacement. Clin.Orthop. 112 :170-179. 2 - Chevalier J., Drouin J.M., Cales B. (1997) Low temperature ageing behavior of zirconia hip joint heads. Bioceramics, Volume 10 Edited by L.Sedel and C.Rey 3- Dambreville A., Rolland Jacob G., Lautridou P. (1996) Cotyles metal-back et usure du polyéthylène. Rec.Chir.Orthop. Suppl.2, 89-90 4 - Dambreville A. (1995) Etude comparative de deux séries de prothèses totales de hanche hydroxyapatite versus titane poreux. Cahiers d’Enseignement de la SOFCOT n° 50 :159-165 5- Dambreville A., P.Lautridou (1996) Les cotyles impactés European Journal of Orthopaedic Surgery and Traumatology 6 :217-222 6 - Derbyshire B., Fisher J., Dowson D., Hardaker C., Brummitt K. (1994) Comparative study of the wear of UHMWPE with zirconia ceramic and stainless steel femoral heads in artificial hip joints. Med.Eng.Phys. 16 : 229-236 7 - GACON G. (1998) A propos des tiges fémorales cimentées en alliage de titane. Maîtrise Orthop. n° 77: 1-24-27 8 - Heindel R., Cales B. (1992) Radioactivity of zirconia ceramics used for femoral heads 4th World Biomaterials congres Berlin 9 - Le Mouël S., Allain J., Pidet O., de Clavière G., Goutallier D. (1997) Premiers résultats alarmants du couple Zyrcon/polyéthylène dans les prothèses totales de hanche Rev.Chir.Orthop. 83 Suppl.II : 44 10 - Le Mouël S., Allain J., Goutallier D. (1998) Analyse actuarielle à 10 ans d’une cohorte de 156 prothèses totales de hanche cimentées à couple de frottement alumine/polyéthylène. Rev.Chir.Orthop. 84:338-345. 11 - Lerouge S., Huk O., Yahia L.H., Witvoet J., Sedel L. (1997) Ceramic-ceramic and metal polyethylene total hip replacements. J.B.J.S. (B) 79 : 135-139