Summary. Sequentially irradiated and annealed UHMWPE hip and knee retrievals showed subsurface in vivo oxidation in both the articular surface and unloaded surfaces, while three of four never-implanted shelf stored liners had oxidation in the bulk. Introduction. Highly cross-linked polyethylene was developed to improve the wear resistance of UHMWPE bearing surfaces in total hip arthroplasty. First generation irradiated and annealed polyethylene showed high oxidation in vivo, largely attributed to only the partial-quenching of free radicals, along with additional radicals generated during terminal gamma sterilization. A second generation, three-step sequential irradiation and annealing method was advanced with the promise of better oxidative stability and improved mechanical properties. We hypothesised that without the complete elimination of free radicals combined with gas plasma sterilization requiring oxygen-permeable packaging, that this second generation material would be prone to shelf-oxidation in addition to in vivo oxidation. Patients & Methods. Fifty surgically-retrieved sequentially irradiated and annealed, gas plasma-sterilised UHWMPE acetabular liners and tibial bearings (X3™, Stryker, Mahwah, NJ), with in vivo durations of 0.5–73 months, were analyzed at their articular surface and an unloaded surface, along with four never implanted acetabular liners. Infrared microscopy was used to evaluate lipid absorption, oxidation (per ASTM F2102-01ε1) and hydroperoxide levels after nitric oxide staining. Gravimetric swelling analysis assessed cross-link density (per ASTM F2214), and crystallinity measurements were performed using differential scanning calorimetry. Results. There was detectable oxidation (OI > 0.1) in 37 of the 50 components with as little as 2 weeks of in vivo service. Maximum oxidation values averaged OI = 0.30 ± 0.30 (range = 0.03–1.59). Oxidation profiles were predominantly characterised by subsurface oxidation peaks approximately 1–2 mm below the surface, in both the articular surface and rim, along with a pattern of embrittlement induced white banding in four and six year liners. Three short in vivo duration liners (0.1–15.5 month) showed oxidation and degradation of material properties throughout the bulk. Three of four never-implanted liners, with up to five years shelf storage, also showed bulk oxidation (Max OI ≤ 1.5), loss of cross-link density and increased crystallinity. Discussion/Conclusion. High levels of detectable oxidation, subsurface oxidation peaks, and white banding were all identified in sequentially irradiated and annealed UHMWPE retrievals with short in vivo durations. These results raise concerns about the long-term clinical performance of these materials. Oxidation measured in shelf-stored, never implanted liners also raises concerns that liners may already be oxidatively compromised before being implanted into patients. Due to gas plasma sterilization methods, these free-radical containing liners are packaged and stored in air, likely resulting in a pre-implantation oxidation effect similar to that historically reported in gamma-in-air sterilised UHMWPE. Longer-term retrievals are needed to better understand the progress of these in vivo changes and whether or not it will compromise the longevity of the implants
