April 1996 Bulletin

Does gamma radiation speed or slow wear?

by Harry A. McKellop, PhD

Harry A. McKellop, PhD, is an associate professor of orthopaedics and biomedical engineering,
director of research for the department of orthopaedics, and director of the
J. Vernon Luck Orthopaedic Research Center at Orthopaedic Hospital, Los Angeles.

Excessive wear of the polyethylene bearing surfaces can substantially shorten the lifespan of prosthetic hips and knees. Wear particles averaging less than a micron in size that accumulate in the periarticular tissues may stimulate an inflammatory foreign body response, aggressive localized bone resorption (osteolysis) and progressive disruption of the implant-bone interface, undermining the fixation of the components.

In a concerted effort to reduce or eliminate the wear of polyethylene in prosthetic joints, a number of laboratories in academic institutions and private industry are investigating the effects of the method of sterilization on long-term wear resistance of ultra high molecular weight (UHMW) polyethylene.

Historically, polyethylene components were sterilized by ethylene oxide gas, but for practical reasons, this was largely supplanted by exposure to gamma radiation, typically at doses ranging from 2.5 to 4.0 Mrads, with the components packaged in air. It has long been recognized that gamma irradiation of polyethylene results in bond scission and the formation of free radicals. If oxygen is present, either mixed with the original polymer powder or having diffused in after fabrication and sterilization, it may react with certain free radicals, lowering the molecular weight of the polymer. This leads to increases in density, stiffness and brittleness, and reductions in the fracture strength and elongation to failure, all of which theoretically, might reduce the wear resistance. Furthermore, due to the presence of long-lived free radicals, oxidation may increase for years after irradiation, sometimes reaching a maximum in a zone located about 0.5 to 2.0 millimeters below the surface of the component.

Improves wear resistance

On the other hand, particularly if little oxygen is present during irradiation, the free radicals may form bonds between carbon atoms on adjacent molecules, cross-linking the polyethylene into a single massive network. There is growing evidence that cross-linking improves the resistance to wear. Since free radicals that have formed cross-links are no longer available for oxidation, the two are competing processes. FTIR measurements of recently gamma-sterilized polyethylene components indicate oxidation to be at a maximum at the free surface, decreasing with depth into the polymer. Conversely, gel content measurements indicate cross-linking to be at a minimum on the surface, increasing with depth into the polymer.

Thus, with recently gamma sterilized UHMW polyethylene components, the wear rate also may vary with depth. For example, in hip simulator studies performed in the author's laboratory, the wear rate of acetabular cups irradiated in air was initially greater than for nonirradiated cups, but decreased over several million cycles (the equivalent of several years' use in a patient) as the wear progressed through the more oxidized surface layer and into the more cross-linked subsurface region. Eventually, the wear rate was slightly lower with the irradiated cups.

In clinical use, the long-term wear resistance of a specific acetabular cup or tibial plateau will depend on numerous factors, including the type of polyethylene from which it was fabricated, the fabrication method (e.g., molding or extrusion), the type of sterilization used, the type of implant (hips and high or low conformity knees), the duration of shelf-storage prior to implantation, duration of use in vivo, and the depth within the polymer to which wear has progressed. The influence of each of these complex and interactive parameters on the long-term wear resistance of polyethylene components is only now being isolated and quantified through clinical radiographic wear measurements, analysis of retrieved implants, and laboratory wear simulations. The following generalities are based on a recent survey of the literature and unpublished data from a number of laboratories.

The wear rate in vivo among acetabular cups of a particular design can vary by a factor of ten or more, for example, from as low as 0.05 mm per year to as much as 0.5 mm or more. However, the wear rate of a specific cup tends to be nearly constant over the years, even though the level of postirradiation oxidation is presumably increasing. In addition, analysis of retrieved cups has shown little correlation between the extent of oxidation and the average rate of wear (i.e., total wear divided by years of use). In the absence of gross third-body abrasive damage, the worn zones of acetabular cups tend to be smooth and glossy, without evidence of pitting or delamination due to macro-fatigue wear. This also is true of relatively conforming, inherently stable knee replacements where, as with a hip prosthesis, the larger contact area leads to lower stress on the surface of the polyethylene.

Extensive pitting

In contrast, with relatively nonconforming, less stable knee replacements, the stresses are greater than with conforming contact and the maximum stress may occur 1 to 2 mm beneath the surface of the polyethylene. The polyethylene components sometimes experience extensive surface pitting and delamination that is strongly correlated with the extent of radiation-induced oxidation. This is particularly true when the oxidation is a maximum in a "white band" below the surface, in the same region as the maximum stress. However, oxidation is not the only contributing factor. The likelihood of pitting and delamination is increased by the presence of fusion defects, which may function as stress concentrations and crack initiation sites, and by the use of a thin polyethylene component, which may substantially increase the stress. On the other hand, the presence of one or more of these factors does not guarantee that fatigue wear and delamination will occur. More research is needed to clarify the relative influence of material properties and design variables on the in vivo performance of total knee replacements.

Need studies

While it is clear that, both for hips and knees, radiation-induced oxidation should be minimized, the optimal method for doing so is presently a subject of controversy. That is, is it preferable to avoid gamma sterilization altogether, in order to avoid even minimal oxidation, or to gamma sterilize the cups in a reduced oxygen environment in order to maximize the benefits of cross-linking? These questions may be addressed by properly-designed laboratory wear simulations with hip and knee prostheses, which can generate the equivalent of five to 10 years of use of a prosthesis in a few months. However, the studies need to be run both on freshly sterilized components and on components that have been artificially aged to model the long-term oxidation in vivo. For example, the oxidation rate may be accelerated by heating the components to 70-80°C in air or in high pressure oxygen chambers for a week or more.

Continuing research is focused on determining the appropriate aging conditions for inducing the same type and amount of oxidation as occurs in a polyethylene component after a specific number of years in vivo.

In a recent hip joint simulators study,1 acetabular cups that were gamma irradiated in air and artificially aged exhibited a several-fold increase in wear rate over cups that were sterilized by gas plasma and aged (presumably, due to the lack of oxidation in the latter). Similarly, in tests in the author's laboratory, cups that were gamma irradiated in low oxygen environments (including vacuum, inert gas, or with an oxygen scavenger present) exhibited from 15 to 30 percent lower wear than cups irradiated in air, presumably due to reduced oxidation and increased cross-linking, and the advantage increased after artificial aging. The same effects have been observed in tests simulating wear in low-conformity knee joints, where pitting and delamination has occurred with air-irradiated and aged polyethylene tibial plateaus, presumably due to the formation of the subsurface brittle band of oxidized polymer, but not with polyethylene that was irradiated in inert gas, thermally-stabilized and then aged.2

The key question that remains is whether, for a given design of hip or knee prosthesis, will it be more advantageous in the long run to sterilize with non-radiation methods such as ethylene oxide or gas plasma to eliminate oxidation, or to irradiate in a reduced oxygen environment to minimize oxidation while retaining the benefits of cross-linking? This question is being addressed in an NIH-sponsored study3 in the author's laboratory in a hip simulator wear test of acetabular cups that were fabricated from a single batch of GUR 4150 HP UHMW polyethylene4 and subsequently modified by four different manufacturers. The modifications included sterilization by gas plasma and by gamma irradiation in several low oxygen environments, one that included post-sterilization thermal stabilization to eliminate free radicals and further cross-link the polymer. The wear rates will be compared before and after artificial aging to the equivalent of five to 10 years in vivo. It will be valuable to perform similar experiments with total knee replacements since, for the reasons summarized above, the relative influence of oxidation and cross-linking might be quite different with some designs of tibial plateaus than with acetabular cups.


  1. M. Sanford, DePuy-DuPont Orthopaedics, Newark, Del., unpublished data.
  2. Blunn G and Bell C: Trans. ORS1996, 482.
  3. NIAMSSD Grant # AR40996-01A2.
  4. For qualified laboratories, samples from this reference batch of UHMW polyethylene are available from:
    Stephen Li, PhD, Biomechanics Laboratory,
    The Hospital for Special Surgery,
    535 East 70th Street,
    New York, N.Y. 10021. Fax (212) 606-1490.

Home Previous Page