Rates of in vivo (arterial) and in vitro biocorrosion for pure magnesium

The development of magnesium-based materials for bioabsorbable stents relies heavily on corrosion testing by immersion in pseudophysiological solutions, where magnesium degrades faster than it does in vivo. The quantitative difference in corrosion kinetics in vitro and in vivo is largely unknown, but, if determined, would help reduce dependence on animal models. In order to create a quantitative in vitro-in vivo correlation based on an accepted measure of corrosion (penetration rate), commercially pure magnesium wires were corroded in vivo in the abdominal aortas of rats for 5-32 days, and in vitro for up to 14 days using Dulbecco's modified eagle medium. Cross-sectioning, scanning electron microscopy, image analysis, a modified penetration rate tailored to degraded wires, and empirical modeling were used to analyze the corroded specimens. In vitro penetration rates were consistently higher than comparable in vivo rates by a factor of 1.2-1.9× (±0.2×). For a sample <20% corroded, an approximate in vitro-in vivo multiplier of 1.3 ± 0.2× was applied, whereas a multiplier of 1.8 ± 0.2× became appropriate when the magnesium specimen was 25-35% degraded.