An investigation into the mechanisms of failure of current Bjork-Shiley cardiac valve prostheses is reported. Two failed valves, one apparently unfailed but defective valve, and one unused valve, were examined by scanning electron microscopy and metallographic section. In the first two valves (removed 12 and 23 months after implantation) fracture was associated with the welds joining the short strut to the valve ring. The fracture surfaces in all cases were heavily faceted and showed branching cracks. Extensive wear had occurred on one fracture surface in the first case, suggesting that one leg of the short strut had failed before the other, though this had been clinically undetectable. The third valve was removed owing to failure of the suturing (24 months after implantation) but one leg of the short strut was found to be completely fractured. The other leg showed extensive cracking and porosity in the weld region. A metallographic section taken through the weld region of the fourth (unused) valve illustrated several sizable defects directly attributable to the welding process. It is suggested that the valves failed by fatigue and that these problems could be overcome if the complete valve cage were machined as a single piece.

No abstract

No abstract