Tribological evaluation.pdf

ELSEVIER

PI1 SO142-9612

(96)

00152-4

Eiomoteriols 18 (1997) 441-447

0 1997 Elsevier Science Limited

014%9612/97/$17.00

Tribological evaluation of oxidized

zirconium using an articular cartilage

counterface: a novel material for

potential use in hemiarthroplasty

A.M. Pate1 and M. Spector

Department of Orthopedic Surgery, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA 02115,

USA; and Rehabilitation Engineering R&D Laboratory, BrocktonlWest Roxbury VA Medical Center, West Roxbury,

MA. USA

Problems with total joint replacement that have surfaced in recent years have made reconsideration of

more conservative hip reconstructive procedures compelling. Moreover, procedures such as hemiar-

throplasty might benefit from newly developed materials that could provide more favourable tribolo-

gical performance when employed as a counterface for articulation with articular cartilage. The

objective of this study was to evaluate the tribology of a new biomaterial, oxidized zirconium, with

articular cartilage in a laboratory test apparatus. Oxidized zirconium components are produced by

oxidizing the zirconium alloy to form a relatively thick (7,um), adherent, abrasion-resistant ceramic

surface. We found that the coefficient of friction of bovine articular cartilage rubbed against oxidized

zirconium was lower than with cobalt-chromium alloy control surfaces, and that there was a trend

toward less wear with oxidized zirconium. A defined layer of degraded tissue was found on the

surface of the articular cartilage specimens, providing some indication of the mechanism of wear.

Results of this study warrant further investigation

of oxidized zirconium as the bearing surface for

hemiarthroplasty.

Keywords: Hemiarthroplasty,

wear, articular cartilage, oxidized zirconium

Received 19 June 1996; accepted 3 September 1996

Concern about the longevity of total joint replacement

has renewed interest in hemiarthroplasty as a treatment

for certain hip disorders. Hemiarthroplasty has been

well established as a practical and beneficial treatment

for elderly patients with femoral neck fractures

because of low morbidity and expedient return to

independence14. Young patients with avascular

necrosis of the femoral head have also been considered

to be good candidates for conservative reconstructive

surgical procedures involving hemiarthroplasty.

However, reports of relatively early failure19234* have

made this indication for the prosthesis questionable.

Causes of failure of hemiarthroplasty prostheses

include pain and the erosion (i.e. wear) of acetabular

articular cartilage and bone. These processes are

probably related in that wear debris can elicit

inflammatory processes that provoke the release of

neurotransmitters associated with pain. Moreover,

innervation of acetabular tissues with nociceptive

nerve fibres7 can be accompanied by an increase in the

amount of released neurotransmitters (e.g., substance

P) that are known to promote inflammations-10

and

degradative processes that might facilitate erosion of

tissue by the prosthesis. This might be reflected, in

part, in the high incidence of protrusion of the hip in

rheumatoid arthritis. Questions remain regarding the

relationship between design features and the erosive

processes underlying failure of hemiarthroplasty.

Design specifications for hemiarthroplasty prostheses

include head size, stem design, method of fixation (i.e.

cemented or non-cemented), unipolar versus bipolar

articulation and material from which the head is made.

Initially, prostheses were available in few head sizes,

leading to incongruence that resulted in higher contact

stresses. In addition, the non-cemented stem designs

were not as canal filling as was later found to be

necessary for mechanical stability. Bipolar devices

were developed to reduce the sliding contact with

tissue. However, despite the fact that both unipolar

and bipolar designs have been employed for many

years, a direct comparison of their performance has

not yet been possible. Unipolar devices have only

recently become available with head sizes that have l-

mm increments in diameter to allow for closer

matching of the head to the acetabular socket.

Moreover, unipolar devices have generally not been

made available

Correspondence to Dr M. Spector, Department of Orthopedic

Surgery, Brigham and Women’s Hospital, 75 Francis St.,

Boston, MA 02115, USA.

with ‘modern’

canal-filling

stem

designs. While direct comparison

of unipolar and

441

Biomaterials 1997,

Vol. 18 No. 5

442

Oxidized zirconium for hemiarthroplasty:

A.M. Pate/ and M. Spector

bipolar devices has not yet been possible, clinical

studies indicate that implementation of bipolar

prostheses has not significantly reduced acetabular

erosion”, while introducing concerns of polyethylene

wear. A recent studyI revealed femoral osteolysis in

58% of 45 primary uncemented bipolar hemiarthro-

plasties in 38 patients followed for an average of 5

years. Eight of these prostheses that were revised

displayed wear of the polyethylene insert rims without

substantial wear of the bearing surface. Histological

examination of the peri-implant tissue revealed

polyethylene particles in macrophages and

multinucleated foreign body giant cells. The authors

concluded that ‘the intrinsic functional mechanism of

the bipolar endoprosthesis leads to persistent cyclic

impingement of the stem neck on the rim of the

bipolar head, producing polyethylene particles and

frequent stem loosening by periprosthetic osteolysis’.

A critical element of the hemiarthroplasty prosthesis

is the material from which the head is produced and

its topography. Previous laboratory studies have

revealed some of the determinants of wear of articular

cartilage when it is articulated with metal counterfaces.

Lipshitz and Glimcher’” developed a laboratory test

apparatus to rub articular cartilage against a stainless

steel counterface. They found that the wear rate of

cartilage increased with roughness of the metal and

that the presence of a third body greatly increased

wear. In these experiments, calcium phosphate

particles were employed as the third bodies in order to

simulate the conditions when bone fragments or

mineral deposits are introduced between the

articulating components. Also of note was the finding

that there was no correlation between the orientation

of the cartilage plug, relative to the direction of sliding,

and wear rate.

Cobalt-chromium alloy is the primary material

employed as the bearing surface for the hemiarthro-

plasty device. However, animal14 as well as the

aforementioned human studies have documented the

degradative changes in articular cartilage associated

with the use of this material for hemiarthroplasty

prostheses. Moreover, comparable findings were

reported in an investigation employing cobalt-

chromium alloy devices to resurface the trochlea of the

dog knee’“. The study documented the degradative

changes in the articular cartilage of the opposing

patella 3 months after implantation of the hemiarthro-

plasty device. The investigation also demonstrated the

increase in wear rate with roughness of the metallic

surface.

In an attempt to develop prostheses with improved

tribological performance, several investigators have

studied the effects of the material of fabrication of the

hemiarthroplasty device on the wear of the opposing

articular cartilage using canine models for hemiarthro-

plasty of the hip. Mendes et al.lfi demonstrated the

unsuitability of high-density polyethylene as the

material of fabrication of the hemiarthroplasty head.

Significant wear of articular cartilage and the high-

density polyethylene was found with these devices.

Investigation of similar devices employed in human

studies yielded similar findings17. Canine studies

employing alumina heads identified some erosion

through the articular cartilage to bone after 2 years’a.

These studies demonstrated, however, that despite the

articulation of the head with bone, there was no

scratching of the ceramic. This finding led the investiga-

tors to recommend ceramic for hemiarthroplasty

devices. A more recent study showed that there was less

wear of articular cartilage with carbon heads than with

cobalt-chromium alloy or titanium alloy devices’“. In

summary, results of previous animal investigations have

suggested that some materials might be superior to

cobalt-chromium alloy for hemiarthroplasty. These

studies have confirmed laboratory experiments that

have shown that the wear rate of cartilage increases with

the roughness of the metal counterface. Benefits have

been demonstrated with the use of scratch-resistant

ceramic materials. Moreover, there has been some

evidence that certain materials, perhaps by virtue of

their lubricity, yield less wear of articular cartilage than

cobalt-chromium alloy.

The objective of this study was to investigate the

tribology of a novel abrasion-resistant material,

oxidized zirconium, when used as the counterface

against articular cartilage. An accelerated wear test

was used in this laboratory screen of the material for

hemiarthroplasty. Oxidized zirconium components are

produced by oxidizing particular zirconium alloys to

form a hard, abrasion-resistant ceramic surface’“.

The ceramic, zirconium oxide (zirconia), has proven

to be of value for decades for a wide range of non-

medical applications, and more recently for the

fabrication of femoral heads for total hip replacement

prostheses. In contrast, the metal, zirconium, and its

alloys owe their development principally to the

nuclear industry, where they have been able to meet

the special needs of a cladding material for pressurized

water vessels”. Zirconium alloys are commended as

nuclear core materials by their small neutron

macroscopic absorption cross-section”“. Advantages of

oxidized zirconium alloys for other applications, such

as the fabrication of components of total joint

replacement prostheses, relate to abrasion resistance

and thickness of the oxide layer. Transmission electron

microscopy

of the oxide scale on certain zirconium

alloys”2-2”

has provided evidence

of the adherent

nature of the oxide scale.

MATERIALS AND METHODS

Oxidized zirconium

Zirconium-2.5niobium alloy was oxidized using time

and temperature conditionszO to form an oxide 7pm in

thickness (Smith & Nephew Orthopaedics, Memphis,

TN, USA). Previous studies have demonstrated how

the thickness of the oxide layer and the size and

orientation of the blue-black monoclinic zirconium

oxide crystallites comprising the oxide can vary with

the oxidation conditions20,22. 23. 26. Discs fabricated

from oxidized zirconium and cast and wrought cobalt-

chromium alloy controls had a highly polished finish

(R, < 0.03 pm).

Laboratory apparatus for an accelerated wear

test

A pin-on-disc

wear apparatus

(Implant

Sciences

Company,

Danvers,

MA, USA) was modified

evaluate the tribological performance

of biomaterials

Biomaterials

1997, Vol. 18 No. 5

Oxidized zirconium for hemiarthroplasty:

A.M. Pate/ and M. Spector

443

with articular cartilage counterfaces. Osteochondral

plugs from calves and adult cows, 7mm in diameter,

served as the pin and the biomaterials as the disc. The

apparatus was set up so that the disc rotated in one

direction, with the pin circumscribing a path with a

radius of 7 mm (to the inner edge of the pin), yielding

22 736 revolutions per km. The test was configured to

maintain articular cartilage in a fluid environment at

room temperature while generating loaded sliding

contact with the discs. The apparatus was instrumented

with a Chatillon 250-g force transducer (Commercial

Scales) to measure the lateral load on the pin for

calculation of the coefficient of friction. A PC-based

data acquisition system was implemented for real-time

measurements of the friction forces.

Normal loads of 1.5 and 2.5 kg were used in this

study. We found that higher load, using the cobalt-

chromium alloy control in our test apparatus, resulted

in exceedingly high frictional forces and frequently in

shear failure of the articular cartilage from its bone

substrate. Assuming a contact area of approximately

0.3cmZ for the slightly convex cartilage specimen, the

contact stresses used in this study were about 5 and

8 kgcrn--’ for the 1.5 and 2.5 kg loads, respectively.

These values are at the low end of the physiological

rangez7, compared to the relatively high stress

(82 kg cm ‘) employed in a previous investigation13.

The articulation was lubricated with sterile, filtered

(triple filtration with 0.1 {trn filter) bovine calf serum

(Hyclone Labs, Inc.). McKellop and Luz8 had

previously reported a build-up of calcium phosphate

on the surface of zirconia balls tested in a hip simulator

with bovine serum as the lubricant. We performed X-

ray photoelectron spectroscopy on discs of oxidized

zirconium and cast cobalt-chromium

anomalous friction measurements due to edge effects.

The calf cartilage differed histologically from the adult

specimens in that the chondrocytes were less columnar

in their arrangement and the collagen less organized.

The histology of adult bovine articular cartilage is

similar to that of human material.

The bone portion of the plug was mounted in

poly(methy1 methacrylate) to facilitate stable fixation

to the wear test apparatus. Plugs were stored at 4°C in

phosphate-buffered

saline for a maximum of 5 days

until tested.

Coefficient of friction

Friction analysis was performed on oxidized

zirconium, cast cobalt-chromium and forged cobalt-

chromium articulated with the calf cartilage. Tests

were performed at speeds of under 0.1 ms-‘; this

speed was in the middle of the physiological range27.

The real-time measurements of the frictional forces

recorded during the tests were divided by the normal

load to obtain the coefficient of friction. These values

were plotted against slide distance. Coefficients of

friction for the materials were then determined by the

asymptotic values of the real-time measurements.

Statistical analysis of the difference in coefficient of

friction between groups was performed

using the

Student t-test.

Wear analysis

alloy specimens

Quantitative analysis of wear involved determination of

the linear wear by measuring the decrease in thickness

of the cartilage resulting from articulation with the test

and control counterface. Before a cartilage plug was

used in the wear test, the thickness of the cartilage was

measured with calipers at four points around the

circumference and then averaged to obtain an estimate

for the thickness of the cartilage at the centre of the

plug. Normal loads of 1.5 and 2.5 kg were applied and

tests were run for a slide distance of 9 km. Immediately

after testing, the plug was stored in 10% neutral

buffered formalin. After 24 h, the plug was sliced in

half, perpendicular to its surface, and the thickness at

the centre of the plug measured. This value was then

subtracted from the initial estimate to obtain the

change in thickness. Statistical analysis of the

difference in wear between groups was performed

using the Student t-test.

After taking the thickness measurement, one-half of the

cartilage was prepared for histological evaluation. The

cartilage was dehydrated in ethanol and embedded in

glycol methacrylate. Three-micrometre-thick sections

were cut with a Reichert-Jung microtome. These sections

were stained with Masson Trichrome for collagen,

Safranin O/Fast Green for proteoglycans, and Haematox-

ylin and Eosin. Histological evaluation of the specimens

served two purposes: it enabled validation of the wear

measurements and it provided an insight into the

mechanism by which the articular cartilage wears.

Three factors affecting the accuracy of the quantita-

tive analysis of linear wear were: (1) the accuracy of

the pretest estimate of the thickness of the cartilage, (2)

the amount the cartilage swelled due to the damage

that the collagen matrix incurred from the sliding

contact, and (3) the change in thickness of the cartilage

due to creep. We used five osteochondral

after testing. We found no indication

of a calcium

precipitate.

Therefore,

we did not employ chemical

additives

(namely,

chelating

agents)

to prevent

calcification in our lubricant.

The pin-on-disc test, with the operating conditions

used in this study, was employed as a screening

method that could be expected to accelerate the wear

of articular cartilage (compared to conditions in viva).

However, without knowledge of the wear mechanisms

associated with hemiarthroplasty prostheses in human

subjects, it was not possible to determine the degree to

which the wear process was accelerated. That the pin-

on-disc apparatus did not closely simulate in vivo

conditions did not preclude its use for the comparison

of the tribological performance of a new material with

the clinical standard (i.e. cobalt-chromium

alloy).

Osteochondral plug preparation

Stifle joints of calves and adult cows were obtained

within 4 h of slaughter (Research 87, Hopkinton, MA,

USA). The surface of the femoral condyles was isolated

and kept moist with a solution of phosphate-buffered

saline. The technique used for preparation of the

osteochondral plugs is described elsewherezg. The

cartilage on the plugs had a diameter of 7mm and

ranged in thickness from 1.0 to 2.5mm for the adult

cartilage and 4 to 7 mm for the calf samples. All of the

plugs were cored from the same region of the femoral

condyles. Osteochondral plugs from this site had a

slightly convex surface. This geometry

minimized

plugs to

Biomaterials 1997, Vol. 18 No. 5

444

Oxidized zirconium for hemiarthroplasty: A.M. Pate/ and M. Spector

evaluate the accuracy of the pretest estimate. Four

circumferential measurements of the thickness of the

cartilage were taken for each plug. The plugs were

then sliced perpendicular to their surfaces and

thickness measurements at the centre of the plugs

were taken. These values were then compared with the

average value of the circumferential measurements to

evaluate the accuracy of the pretest estimate. There

was no significant difference. The degree to which

swelling and creep could affect the determination of

linear wear was assessed in experiments using control

samples of calf cartilagezg.

After wear testing, the test and control counterfaces

were gently rinsed with water to remove any residue

of the bovine serum lubricant. Samples were sputter

coated with gold. The counterfaces were then evaluated

using conventional scanning electron microscopy.

wear layer from the underlying cartilage, which

appeared normal with respect to its staining for collagen

and proteoglycan. The thickness, stain characteristics

and consistency of the wear layer produced by articula-

tion with the two types of biomaterials were comparable.

On some specimens, the wear layer was non-uniform in

thickness, accumulating on one side of the plug (Figure

3), apparently related to the direction of rotation of the

disc. When stained with Safranin O/Fast Green, the

RESULTS

Swelling of specimens of cartilage due to the damaged

collagen, produced by removing the superficial layer

of control specimens of cartilage with a scalpel, was

found to be 0.1 f 0.1 mm (mean* s.d.; n = 6). The

effect of creep, under the constant load of 2.5 kg used

in the wear experiment, was to decrease the thickness

of the cartilage 0.1 f O.Omm (n = 4). Based on these

findings, it was presumed that there was no remarkable

net effect of creep and swelling on the linear wear

measurements.

The linear wear of the adult bovine cartilage

articulated with the oxidized zirconium was

approximately 30% less than the wear of cartilage tested

with the cast cobalt-chromium alloy counterface

(0.34 f 0.13 mm versus 0.47 f 0.20 mm, mean f sd.;

Figure 1). This difference was, however, not statistically

significant (P > 0.1). A distinct layer of material, which

presumably was degraded cartilage, was found on the

surface of all of the osteochondral plugs after articulation

with the test and control materials. This ‘wear layer’ was

generally uniform in histological appearance and

thickness (approximately 50-75 pm) across the surface

(Figure 2). There was a distinct border separating the

T 1

0.5

0.4 1

T 1

Linear

Wear

(mm)

0.3

0.2

0.1

OJ---

ii&z

Ox kc

n=6

Figure 2 a, Micrograph of adult bovine cartilage stained

with haematoxylin and eosin (H&E). b, Section of adult

bovine cartilage that has articulated against oxidized

zirconium for 9 km. A 50-pm-thick wear layer is apparent.

H&E staining. c, Adult bovine cartilage that has articulated

against cast cobalt-chromium

Figure 1 Graph showing the linear wear of articular

cartilage with oxidized zirconium and cast cobalt-

chromium counterfaces. The heights of the columns

correspond to the mean values and the vertical bars the

standard errors of the mean.

for 9 km. H&E staining.

Biomaterials 1997, Vol. 18 No. 5

Oxidized zirconium for hemiarthroolasty:

A.M. Pate/ and M. Spector

445

0.0Z5

r 1

Coefficient

0.015

Friction

0.01

Castr Forged CoCr

n=4

Ox Zirc

II=4

n=4

Figure 3 Adult bovine cartilage that articulated with cast

cobalt-chromium for 9 km. A wear layer (200-250pm thick)

accumulated on one side of the plug, apparently associated

with the direction of rotation of the disc (i.e. from left to

right). H&E staining.

Figure 4 Comparison of the coefficients of friction for

articular cartilage articulated with oxidized zirconium and

cobalt-chromium alloy specimens. The heights of the

columns correspond to the mean values and the vertical

bars the standard errors of the mean.

degraded cartilage appeared to be deficient in

proteoglycan. This loss of proteoglycan was consistent

with a failure mechanism found in an animal model of

hemiarthroplasty14.

One of four of the calf cartilage plugs articulating

with the oxidized zirconium and three of four

samples in the groups articulating with the cast and

forged cobalt-chromium samples failed prior to

completion of the test. The articular cartilage sheared

from the subchondral bone. This appeared to be

related to elevations in the frictional forcezg. The

wear for the remaining cartilage samples tested

against the oxidized zirconium was 0.5 f 0.1 mm

(meanfs.d.; n = 3) compared to 0.8mm (n = 1) for

the plugs with the forged counterface and 1 mm for

the cartilage

with the cast cobalt-chromium

specimenZg.

Using the 1.5 kg load in tests with the calf cartilage,

the coefficient of friction for the oxidized zirconium

was 0.015 f 0.002 (meanfs.d.; R = 4; Figure 4). This

was significantly less (P < 0.0001) than the values

for the cast (0.021 f0.003; n = 4) and the forged

(0.023 f 0.003; n = 4) cobalt-chromium specimens.

Results from tests of adult bovine cartilage using the

2.5 kg load were variable. In half of the tests of the

oxidized zirconium and cast cobalt-chromium

counterfaces, the coefficients of friction were initially

approximately 0.01, rising to more than 0.17; there

was an abrupt increase in frictional force that

exceeded the force transducer’s capacity. The other

half of the tests started with frictional forces that

were out of range of the transducer. Only one of the

tests with cast cobalt-chromium alloy returned to

range, whereas two of the oxidized zirconium

returned.

Wear tracks were apparent on all of the counterfaces.

On certain discs, there was a clear demarcation

between regions that appeared to have a transfer layer

of degraded cartilage and the featureless regions where

articulation had not occurred. On other counterfaces,

there was a gradual decrease of the amount of transfer

film as distance from the centre of the wear track

increased (Figure 5). There were no marked differences

in structure of the transfer layer on the test and control

counterfaces.

Figure 5 a, Scanning electron micrograph of the centre of

a wear track formed on a disc of cast cobalt-chromium

alloy, displaying a transfer layer of articular cartilage from

the osteochondral plug. b, Scanning electron micrograph

showing the transfer layer on a cast cobalt-chromium

alloy

disc at approximately

10pm from the centre of a wear track.

DISCUSSION

Results of this study showed that the coefficient of

friction of articular cartilage rubbed against an oxidized

zirconium counterface was significantly lower than

Biomaterials

1997, Vol. 18 No. 5

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: