This is an enhanced PDF from The Journal of Bone and Joint Surgery
The PDF of the article you requested follows this cover page.
Post-irradiation aging of ultra-high molecular weight polyethylene
CM Rimnac, RW Klein, F Betts and TM Wright
J Bone Joint Surg Am. 1994;76:1052-1056.
This information is current as of January 8, 2010
Reprints and Permissions Click here to order reprints or request permission to use material from this
article, or locate the article citation on jbjs.org and click on the [Reprints and
Publisher Information The Journal of Bone and Joint Surgery
20 Pickering Street, Needham, MA 02492-3157
(()p right I ‘194 h iI,t J,n,r,,aI of Ho,, 1011 Jo,iti .S,rerv. Ii,corjur,I
Post-Irradiation Aging of
Ultra-High Molecular Weight Polyethylene”t
BY CLARE M. RIMNAC. PH.D4. ROBERT W. KLEIN. B.S.t. FOSIER IJEITS. PFI.D.:.
AND TIMOTHY M. WRIGHT. PH.D4. NEW YORK. N.Y.
IllVeStig(ItU)Il J)erforlned at the De/)artl?lellt ()fBiol?zec/lallics. Tue IIosjitalfor Special Siir,t,’erv.
The New York Hospital-Cornell Univer.suv Medical (‘olle,ge. Nett York (it
ABSTRACT: A study was performed to determine the the life span of joint replacements. Excessive wear can
time-course of oxidative degradation and the extent to lead to gross mechanical failure, in the form of fracture
which the degradation proceeded through the bulk of and dissociation”’2’, and to loosening caused by a bio-
ultra-high molecular weight polyethylene joint compo- logical reaction to particulate polyethylene debris’4’7. It
nents that had been irradiated and stored on a shelf. has been shown that stresses that occurred in polyeth-
Standardized cylindrical samples, taken from a single ylene components were consistent with fracture mech-
batch of extruded polyethylene, were cleaned, packaged, anisms that lead to wear damage. The magnitudes and
and sterilized according to protocols used for commen- distributions of these stresses were a function both of
cial joint-replacement components. After sterilization, the geometry of the component and of the elastic mod-
the samples were stored in the packages for time-periods ulus of the polyethylene2. Alterations, such as those
of one day to more than one year. At each interval resulting from oxidative degradation. in physical. chem-
studied, thin sections were cut as a function of depth into ical, and, hence. mechanical properties (including elastic
the bulk of the sample and were used to determine the modulus) could. therefore, adversely affect the wear re-
density and the infrared spectra. Marked alterations in sistance of polyethylene components.
the density and the infrared spectra consistent with con- Polyethylene components sterilized with gamma ra-
tinuing oxidative degradation occurred throughout the diation are known to undergo oxidative degradation.
year of storage on the shelf. The alterations were most both while being stored on a shelf and after implanta-
severe near the surface of the samples. tion’7”2”. Retrieved components and stored compo-
CLINICAL RELEVANCE: The results ofthis study sup- nents both demonstrated changes in density (a physical
port the hypothesis that sterilization of polyethylene property) and changes in infrared spectra (a chemical
joint components with gamma irradiation followed by property) consistent with oxidative degradation. While
storage on a shelf contributes to oxidative degradation. these changes occurred for a considerable depth into
The degradative changes continue with time and occur the retrieved components’, the magnitude of the deg-
not only on the surface but also through the bulk of radative changes was unknown because the initial prop-
the component. Alterations in the material properties erties of the polyethylene from which the components
associated with the degradation increase the stresses on were fabricated was not known. While degradative
the surface of and within the polyethylene total joint- changes were documented for stored polyethylene sam-
replacement components once these components have ples. the changes were not measured as a function of
been implanted and are subjected to loading in vivo. depth from the surface of the sample.
Thus, the wear resistance of a polyethylene component Alterations in material properties caused by degra-
that has been stored on a shelf for an extended time may dation of polyethylene can he expected to affect the
be diminished before the component is implanted. stresses associated with wear damage observed on and
within the components. To assess the magnitude of
Wear damage to ultra-high molecular weight poly- the degradative effect. the changes in physical and
ethylene (hereafter referred to as polyethylene) total chemical properties as a function of both time after
joint components is a serious clinical problem that limits irradiation and location within the component must be
determined. Therefore, an in vitro experiment was con-
*No benefits in any form have been received or will be received ducted in which standardized polyethylene samples, ap-
from a commercial party related directly or indirectly to the subject proximately the size of a condyle from a total knee tibial
of this article. Funds were received in total or partial support of component. were irradiated and were then analyzed for
the research or clinical study presented in this article. The funding
sources were National Institutes of Health Grants AR40191 and
changes in density and infrared spectra as functions of
AR01876 and the Clark Foundation. the time after irradiation and the depth below the sun-
tRead in part at the Annual Meeting of the Society for Bioma- face of the hulk samples. Density was measured because
terials. Birmingham. Alabama. May 2. 1993.
increased density has been associated with increased
lDepartment of Biomechanics. iTie Hospital for Special Sur-
gery, 535 East 70th Street. New York. N.Y. 10021. elastic modulus” and, thus, can he used to assess the
1052 [TIE JOURNAL OF BONE AND JOINT SURGERY
POST.IRRADIATION AGING OF ULTRA-HIGH MOLECULAR WEIGHT POLYETHYLENE 1053
on a shelf in laboratory air at room temperature for one
of seven time-periods (one, eight, twenty-two, thirty-
eight, sixty-eight, 120, and 378 days after sterilization).
0.945 The time-periods were chosen so that most of the meas-
urements were made during the first two months, when
more rapid degnadative changes were expected. After
storage, the samples were frozen at -20 degrees Celsius
.) to minimize additional degradation until they could
be analyzed. Standard principles of first-order kinetics
C’) suggest that such a reduction in temperature slows
w degnadative reactions by more than an order of mag-
nitude. Additional, non-irradiated samples from the
same lot were prepared at the start of the experiment.
These samples were also kept frozen until they could be
A sample size of two was chosen for each time-period
on the basis of previous experience with density and
infrared measurements on polyethylene material’523. For
0 1 2 3 example. thirty-six density measurements made on sam-
DEPTH FROM SURFACE (mm) ples taken at equally spaced distances throughout a nec-
tangular (sixty-three by seventy-six by 305-millimeter)
extruded rod had a mean value of 0.933 gram per cubic
Graph of density plotted against depth below the surface of the
disk. Two disks were evaluated at each time-period. centimeter (standard deviation, 0.0004 gram per cubic
centimeter) with a range of 0.932 to 0934 gram per cubic
severity of the effect of degradative changes on wear centimeter.
resistance. Analysis with infrared spectroscopy was used For each time-period, density as a function of depth
to confirm the presence of chemical products arising from the surface was determined for each disk. A cylin-
from oxidative degradation. dnical core (4.5 millimeters in diameter) was cut from
the center of each disk and was cut into approximately
Materials and Methods 0.2-millimeter-thick sections with a microtome. The sec-
Cylindrical disks, 25.4 millimeters in diameter and
12.5 millimeters thick, were machined from one lot of
polyethylene extruded nod (GUR 415 resin; Hoechst!
Celanese. Houston. Texas). The initial properties pro-
0 0.5 mm DEPTH
vided by the vendor (Westlake Plastics, Lenni, Penn-
sylvania) were a minimum molecular weight of three V 1.0 mm DEPTH
million, a density of 0.933 gram per cubic centimeter, a 0 2.0 mm DEPTH
yield stress of twenty-two megapascals, an ultimate ten- 0- -‘
sile stress of forty-four megapascals, and an elongation 0
thickness was chosen to be approxi-
mately the same as that of a typical polyethylene im-
plant. The same machining techniques (cutting tools, z p. ..0’-
tool speeds, and feed rates) used to machine implant 0
,l ,,, ,,
components were used to fabricate the samples. GUR , .“ II
415 resin was chosen because it is used widely by man- I I
I / I
ufacturens of commercial implants. .....-d” Q--Q--13
All disks were cleaned, sealed in plastic pouches. ; ; ;=:::
and packaged in pairs. The packaged samples were sten-
ilized with gamma irradiation in a single batch by a 0.935
1 10 100 1000
commercial vendor (Isomedix Operations. Whippany,
New Jersey). The packaging and sterilizing procedures TIME (DAYS)
were those used in the manufacture of commercial im- Fio. 2
plants. The estimated dose of radiation ranged from Graph of density plotted against time after irradiation for depths
28,300 to 42,000 gray, as determined from dosimetnic of 0.5, 1.0, and 2.0 millimeters below the surface. Each point is the
estimated density determined from linear interpolation between
measurements performed by the vendor. adjacent density measurements spanning each of the three depths
A package containing a pain of samples was stored and represents the average of the results from the two disks.
VOL. 76.A, NO. 7. JULY 1994
1054 C. M. RIMNAC. R. w. KLEIN, FOSTER BETTS, AND T. M. WRIGHT
0.936 gram pen cubic centimeter, on approximately 0.3
DAY 8 per cent. There was a greater increase in density near
I- 3 DAY 22 the surface of the sample than within the bulk. from
DAY 38 approximately 0.933 to 0.938 gram pen cubic centimeter,
w DAY 120 or approximately 0.5 per cent.
In general, density continued to increase with the
time after the irradiation. By day 120, the degradative
0 effects were no longer concentrated near the surface but
-J extended well into the bulk of the samples. For example,
z the density from the surface to a depth of one-half mil-
0 limeter was constant, at approximately 0.942 gram pen
cubic centimeter, after 120 days of storage. The density
0 was still greater (to a depth of two millimeters) than the
0 density of 0.938 gram pen cubic centimeter at one day
w after the irradiation. The rate of increase in density was
a function of depth. For example. at a depth of one-half
millimeter, the rate of increase in density with time was
greater than that at one on two millimeters below the
0 1 2 3
surface (Fig. 2).
DEPTH FROM SURFACE (mm) In the non-irradiated samples, no appreciable keto-
FIG. 3 carbonyl peak was noted at any depth (Fig. 3). A peak
Graph of ketocarbonyl peak height plotted against depth below was observed, however, in all of the irradiated samples
the surface of the disk. Two disks were evaluated at each time-period. that were examined with infrared spectnoscopic analysis.
There was no detectable absorption in the carbonyl region for the
As with density, the peak height of the ketocarbonyl was
non-irradiated material. The non-irradiated material is plotted at the
detection limit (0.1 centimeter) for peak height measured at approx- more pronounced near the surface of the disk. In gen-
imately 1720 centimeters-’. eral, peak height continued to increase with time. By day
120, elevated peak heights were found well below the
tions were obtained beginning near the surface and con- surface, to a depth of two millimeters.
tinuing to a depth of about three millimeters below
the surface. The actual thickness of the section was de- Discussion
termined with a micrometer. Section thicknesses were Sterilization with gamma irradiation, conducted in
added to obtain cumulative section location with re- an ambient oxygen environment, has been reported to
spect to the surface of the disk. The density of each result in cross-linking, chain scission, and oxidative
section was determined with a gradient column of iso- degradation of polyethylene9’6. Cross-linking is sup-
propanol and water’. pressed and chain scission and oxidative degradation
Infrared spectroscopic analysis was conducted in dominate when polyethylene is irradiated in the pres-
transmission mode for selected time-periods with use ence of oxygen22. Other investigators have shown that
of a spectrophotometer (model 1330; Perkin-Elmer, post-irradiation effects continue with time. Bhateja et
Danbury, Connecticut) on the same sections used for al. exposed polyethylene to lange doses (as high as
evaluation of the density. The chamber of the spec- 1,200,000 gray) of beta radiation and followed the sam-
trophotometer, with the section in place, was purged ples for 112 months. Both the heat of fusion (related
with dry air, free of carbon dioxide, before collection of to the amount of crystallinity) and the melting tem-
the spectra. The peak height at an absorbency of ap- penature continued to change throughout the test. Bin-
proximately 1720 centimeterst (associated with C=O kinshaw et al.4 exposed polyethylene to doses of as
ketocarbonyl groups in the polyethylene) was meas- much as 500,000 gray of gamma radiation and followed
ured. A baseline was drawn on the transmission spec- the samples for 335 days. They also found that crystal-
trum, and the peak height was measured from the linity and mechanical properties continued to change
baseline. An increase in peak height reflects an increase throughout that time.
in the concentration of C=O ketocarbonyl groups in the Streicher exposed polyethylene to 6000 to 25,000
polyethylene. gray of gamma radiation and stored the polyethylene in
air. Changes in the physical, chemical, and mechanical
properties due to the irradiation occurred only for as
The density of the non-irradiated material was con- long as three months. However, these tests were con-
stant into the bulk of the disk (Fig. 1). Sterilization with ducted on thin films (300 micrometers thick) for which
gamma radiation had an immediate effect on density, the changes due to oxidative degradation would be ex-
even at one day after the sterilization, when the density pected to saturate in a relatively short time. As dem-
within the bulk increased from approximately 0.933 to onstrated in the present study. even small doses of
THE JOURNAL OF BONE AND JOINT SURGERY
Posr-IRRADIATI0N AGING OF ULTRAHIGH MOLECULAR WEIGHT POLYETHYLENE I 055
radiation, such as are used commonly for sterilization, is irradiated in an inert environment, cross-linking,
are sufficient to initiate degradative changes that con- rather than chain scission and oxidative degradation,
tinue to occur for at least one year in bulk samples of dominates initially222. The mechanical properties of
polyethylene. In this study. degradative changes oc- polyethylene could be substantially affected, depending
curned as deep as three millimeters into the samples on the extent of cross-linking. Additional studies to
(Figs. 1 and 3) at one year after the irradiation. Thus. a measure the effect of oxidative degradation on cross-
polyethylene joint component that is six millimeters linked polyethylene must be performed before it can be
thick will undergo degnadative changes through its assumed that gamma irradiation in an inert environment
thickness before implantation if the time from steniliza- would be beneficial.
tion to implantation is one year or more. Degradation would be expected to affect the wear
Carhonyl groups are formed by the reaction of ox- resistance of polyethylene adversely, although we know
ygen with long-lived free radicals that are produced of no direct data from wear tests or from clinical stud-
during irradiation”’. At temperatures well below the ies. The effect of lange doses of radiation (more than
melting point of the polyethylene. the lives of these 200,000 gray) on the wear behavior has been studied’3’5,
radicals are measured in years”. The breakdown of but the polyethylene was tested shortly after the stenil-
trapped metastable penoxy on hydroperoxy species im- ization. The changes in density and infrared spectra that
mediately after irradiation may provide one oxygen were measured in the present study were accompanied
source for polyethylene degradation5, but diffusion of by alterations in mechanical properties that continued
oxygen is required for continued degradation. It is not to change dramatically with the time after the inradia-
surprising. therefore, that, consistent with the results of tion. The magnitude of the density changes measured in
previous studies”, the greatest changes in density and the present study (for example. 0.932 to 0.942 gram pen
infrared spectra occurred near the surface, where oxy- cubic centimeter) is equated with a 35 pen cent increase
gen was most plentiful and more oxygen would be cx- in the elastic modulus of polyethylene on the basis of an
pected to he absorbed. empirically derived relationship’. Such an increase in
Sterilization of polyethylene with radiation in the elastic modulus near the articulating surface of poly-
absence of oxygen might diminish or eliminate oxidative ethylene joint components would cause substantial in-
degradation. Sterilization in nitrogen’2’ has been shown creases in the stresses associated with wean damage
to diminish oxidative degradation during the steniliza- occurring on and near these surfaces2.
tion process but, on subsequent exposure to oxygen,
N,, ihe authors thank N. Khanna and J. Weintraub for their assistance with the
degradation resumed2”. Furthermore,when polyethylene experimental measurements.
1. American Society for Testing and Materials: Standard DI505: density of plastics by the density-gradient technique. In Annual Book of
ASTM Sia,tdard.s, Part 8.01, pp. 458-463. Philadelphia, American Society for Testing and Materials. l99().
2. Bartel, D. L.; Bicknell, V. L.; and Wright, T. M.: The effect of conformity. thickness, and material on stresses in ultra-high molecular
weight components for total joint replacement.]. Bone and]oint Surg., 68-A: 1041-1051. Sept. 1986.
3. Bhateja, S. K.; Andrews, E. H.; and Yarbrough, S. M.: Radiation induced crystallinity in linear polyethylenes: long term aging effects.
Polv,ner ]., 2 1 : 739-750, 1989.
4. Birkinshaw, C.; Buggy, M.; and Daly, S.: Mechanism of ageing in irradiated polymers. Polvnzer Degrad. and Stabil., 22: 285-294, 1988.
5. Birkinshaw, C.; Buggy M.; and Daly. S.: The effect of y radiation on the physical structure and mechanical properties of ultrahigh molec-
ular weight polyethylene.]. App!. PolvinerSci., 38: 1967-1973, 1989.
6. Eyerer, P., and Ke, Y. C.: Property changes of UHMW-polyethylene hip cup endoprostheses during implantation. ]. Bio,ned. Mater. Res.,
1$: I 137-I 151. 1984.
7. Grood, E. S.; Shastri, R.; and Hopson, C. N.: Analysis of retrieved implants: crystallinity changes in ultrahigh molecular weight poly-
ethylene. J. Bionied. Mater. Rev. 16: 399-405, 1982.
8. Jahan, M. S.; Wang, C.; Schwartz, C.; and Davidson, J. A.: Combined chemical and mechanical effects on free radicals in UHMWPE
joints during implantation. ]. homed. Mater. Res., 25: 1005-1017. 1991.
9. Kurth, M.; Eyerer. P.; and Cui, D.: Effects of radiation sterilization on UHMW-polyethylene. ANTEC87, Conf Proc., pp. 1193-
10. Kurtz, S. M.; Rimnac, C. M.; Li, S.; and Bartel, D. L.: A bilinear material model for UHMWPE in total joint replacements. Trans. Ortliop.
Res. Soc.. 19: 289. 1994.
1 1. Lombardi, A. V., Jr.; Engh, C. A.; VoIz, R. C.; Albrigo, J. L.; and Brainard, B. J.: Fracture/dissociation of the polyethylene in metal-
hacked patellar components in total knee arthroplasty. ]. Bone and ]oint Surg.. 70-A: 675-679. June 1988.
12. Maksimov, V. L.; Agnivtseva. T. C.; Khaikin, S. V.; and Pukshanskii, M. D.: Features of post-radiation oxidation of trapped radicals in
high and low density polyethylene. Polymer Sci. USSR. 28: 120-124, 1986.
13. Oonishi, H.; Takayama, Y.; and Tsuji, E.: Improvement of polyethylene by irradiation in artificial joints. Radial. Phys. and C/ie,n., 39:
I 4. Peters, P. C.; Engh. C. A.; Dwyer. K. A.; and Vinh, T. N.: Osteolysis after total knee arthroplasty without cement. ]. Bone and ]oint Surg.,
74-A: 864-876, July 1992.
15. Rimnac, C. M.; Wright. T. M.; Klein, R. W.; Betts, F.; and Schapiro, E.: Characterization of material properties of ultrahigh molecular
weight polyethylene before and after implantation. Trans. Soc. Biomater. Implant Retrieval Svmpos., 15: 16, 1992.
16. Roe, R. J.; Crood, E. S.; Shastri, R.; Gosselin, C. A.; and Noyes, F. R.: Effect of radiation sterilization and aging on ultrahigh molecular
weight polyethylene.]. homed. Mater. Res., 15: 209-230, 1981.
VOL.. 76.A, NO. 7. JULY 1994
1056 c. M. RIMNAC, R. W. KLEIN, FOSTER BETTS, AND T. M. WRIGHT
17. Santavirta, S.; Konttinen, Y. T.; Hoikka, V.; and Eskola, A.: Immunopathological response to loose cementless acetabular components.
]. Boneand]ointSurg., 73-B(1):38-42, 1991.
18. Shen, C., and Dumbleton, J. H.: The friction and wear behavior of irradiated very high molecular weight polyethylene. Wear, 30: 349-
19. Shinde, A., and Salovey, R.: Irradiation of ultrahigh-molecular-weight polyethylene. ]. Polymer Sci., 23: 1681-1689. 1985.
20. Streicher, R. M.: Influence of ionizing irradiation in air and nitrogen for sterilization of surgical grade polyethylene for implants. Radiat.
Phys. Client., 31:693-698, 1988.
21. Stulberg, S. D.; Stulberg, B. N.; Hamati, Y.; and Tsao, A.: Failure mechanisms of metal-backed patellar components. Cliii. Orthop., 236:
22. Wiesner, L.: Effects of radiation on polyethylene and other polyolefins in the presence of oxygen. Radiat. Phys. C/tern., 37: 77-81, 1991.
23. Wright, T. M.; Rimnac, C. M.; Stulberg, S. D.; Mintz, L.; Tsao, A. K.; Klein, R. W.; and McCrae, C.: Wear of polyethylene in total joint
replacements. Observations from retrieved PCA knee implants. Clin. Orthop., 276: 126-134, 1992.
THE JOURNAL OF BONE AND JOINT SURGERY