In Vivo Lipid Absorption in Highly Cross-linked UHMWPE Acetabular Liners

Introduction Radiation cross-linking was developed in order to improve the wear resistance of UHMWPE in total joint replacements [1]. Oxidationinducing free radicals formed during the irradiation process are eliminated through melting [2]. Without free radicals present, irradiated and melted UHMWPE has been considered oxidatively stable. However, with minor oxidation being observed in this material after 10 years in vivo [3], other oxidation mechanisms are now under investigation. Additional studies have reported high levels of oxidation in shelf-aged surgical retrievals [4] and shelf-aged simulator-tested acetabular liners [5]. Alternative destabilizing mechanisms such as lipid-induced oxidation and cyclic loading are under investigation. In the absence of residual free radicals, it has been suggested that irradiated and melted UHMWPE could be adversely affected by in vivo conditions such as exposure to lipids. Lipids, which are readily absorbed by polyethylene in the body [6], could oxidize through a cascade of events very similar to that present in UHMWPE oxidation. The peroxidation cascade produces free radicals that can attack polyethylene and form free radicals on the polyethylene backbone and result in oxidation. The rate and amount of lipid absorption into crosslinked UHMWPEs in vivo has not been studied before. We investigated a series of surgically explanted acetabular liners using infra-red spectroscopy to quantify the lipids by measuring the carbonyl content. Materials and Method Ten retrieved irradiated and melted acetabular liners with in vivo durations ranging from 1-42 months were received by our lab in vacuum sealed packaging, having been frozen at -20°C from the time of explantation to prevent ex vivo oxidation. All liners were MarathonTM (DePuy, Warsaw, IN), which are gamma irradiated to 50kGy and subsequently melted, followed by machining and gas plasma sterilization. Thin sections were cut from a cross-section of the high load articular surface of each retrieval (150 μm) using a microtome. Spectra were collected by Fourier Transform Infrared Microscopy (FTIR; BioRad FTS155/UMA500, Natick MA) as a function of depth from articular surface to backside. The thin sections were then boiled in hexane for 16 hours to remove absorbed estrified fatty acids [7] and dried in vacuum. FTIR spectra were collected again after hexane extraction. Carbonyl indices were calculated from spectra, both before and after hexane extraction, as the ratio of the areas under 1680 cm – 1780 cm to the absorbance over 1330 cm – 1390 cm, per ASTM F210201. Correlation was calculated using a logarithmic regression. Significance was determined using an unpaired, two-tailed student’s ttest. Results and Discussion UHMWPE absorbs lipids from the synovial fluid in vivo, including cholesterol, hexaand octadecanoic esters of cholesterol and squalene [6]. Absorbance of these components is measured through FTIR, by quantifying the carbonyl-containing esters at 1740 cm absorbance peak. This same peak is used to identify carbonyl groups found in oxidation products. We were able to differentiate between absorbed lipids and oxidation products by hexane extracting the esterified fatty acids [7]. Pre-hexane extraction carbonyl levels showed an average lipid absorption within the first 3mm of 0.242 ± 0.140 in the articular surface and 0.133 ± 0.082 in the rim. After extraction, oxidation levels (carbonyl indices) were negligible (0.003 ± 0.006). Lipid absorption occurred through a diffusive process. Maximum lipid absorption was nearly constant in the articular surface (0.491 ± 0.062) regardless of in vivo duration (Fig 1). Lipid penetration depth, however, increased with increasing in vivo duration (Fig 2). Regression lines were fit to the penetration depth for both the articular surface (R = 0.95 with a power regression) and rim (R = 0.91 with a linear regression). Absorption at the articular surface was both deeper and more rapid than absorption at the rim. This is likely due to the cyclic loading seen by the articular surface. At longer in vivo durations the depth of lipid diffusion reached 3mm in the articular surface (Fig 1) and 1.5mm in the rim.