Abstract
Introduction
Orthopedic implants are subject to wear and release ultra-high molecular weight polyethylene (UHMWPE) debris. Analysis of UHMWPE wear particles is critical in determining the safety and effectiveness of novel orthopedic implants. Complete digestion of periprosthetic tissue and wear fluid is necessary to ensure accurate morphological and quantitative particle analysis. Acid digestion methods are more effective than enzymatic and base digestion approaches [Baxter+ 2009]. However, optimal digestion times, quantity, and type of acid are unclear for particle isolation. In addition, imaging and analysis techniques are critical to ensure accurate reporting of particle characteristics. Here, we 1) compared the efficacy of three acid-based digestion methods in isolating particles from a) bovine serum and b) animal/human tissue, and 2) analyzed the effects of imaging location on particle quantity/morphology results.
Methods
1a) UHMWPE (GUR 150) particles were generated by Mode I knee wear testing for 1 million cycles in bovine serum. Serum was digested in one of four solutions: 12.2M HCl, 15.8M HNO3, a 1:1 volume ratio of HNO3:HCl (aqua regia), or filtered H2O (control). The serum:solution volume ratio was 1:5 [Niedzwiecki+ 2001, ISO 17853:2011]. Digestion occurred for 60min on a stir plate at 60°C. Each digest was combined with MeOH at a 1:5 digest:MeOH volume ratio and filtered using a 100 nm polycarbonate membrane. The particle-containing membranes were imaged (12 images/membrane) using scanning electron microscopy (SEM) to determine particle characteristics, including quantity, equivalent circular diameter (ECD) and aspect ratio (AR). 1b) Based on 1a, HNO3 was used to digest porcine and human tissue at concentrations of 1:40, 1:60, or 1:80 tissue:HNO3 volume ratios for either 1, 12, or 24 hours, followed by SEM analysis. 2) Particle characteristics were compared at nine locations (20 images/location) across a particle-containing membrane to determine the effects of imaging location.
Results
1a) HNO3 and aqua regia methods successfully digested the bovine serum, whereas the HCl and H2O methods were unsuccessful (Fig.1A). Comparing HNO3 and aqua regia groups, particle characteristics and ECD frequency distribution were nearly identical (Fig.1B). 1b) Nitric acid did not fully digest porcine or human tissues. 2) Similar particle characteristics were observed in all nine locations analyzed across the polycarbonate membrane. The particle quantity, ECD, and AR for a representative center vs. intermediate location were 808 vs. 780 particles, 0.33±0.28 vs. 0.35±0.29 µm, and 1.57±0.56 vs. 1.51±0.4, respectively (Fig.2).
Conclusions
Nitric acid and aqua regia are capable of digesting bovine serum using low quantities of acid for short duration, allowing precise analysis of UHMWPE particle debris from orthopedic implants. However, further optimization of digestion techniques for animal/human tissue is warranted. In addition, an accurate representation of particle distribution can be achieved without analyzing hundreds of images, because membrane location does not strongly influence particle results. Finally, ASTM F1877-16 – Standard Practice for Characterization of Particles – could benefit from adding software-based automated particle characterization as an optional method. An automated approach that uses k-means clustering image segmentation to identify particles and computer vision tools to extract relevant morphological features is under development and validation.
