Revised microcalcification hypothesis for fibrous cap rupture in human coronary arteries.

Using 2.1-µm high-resolution microcomputed tomography, we have examined the spatial distribution, clustering, and shape of nearly 35,000 microcalcifications (µCalcs) ≥ 5 µm in the fibrous caps of 22 nonruptured human atherosclerotic plaques. The vast majority of these µCalcs were <15 µm and invisible at the previously used 6.7-µm resolution. A greatly simplified 3D finite element analysis has made it possible to quickly analyze which of these thousands of minute inclusions are potentially dangerous. We show that the enhancement of the local tissue stress caused by particle clustering increases rapidly for gap between particle pairs (h)/particle diameter (D) < 0.4 if particles are oriented along the tensile axis of the cap. Of the thousands of µCalcs observed, there were 193 particle pairs with h/D ≤ 2 (tissue stress factor > 2), but only 3 of these pairs had h/D ≤ 0.4, where the local tissue stress could increase a factor > 5. Using nondecalcified histology, we also show that nearly all caps have µCalcs between 0.5 and 5 µm and that the µCalcs ≥ 5 µm observed in high-resolution microcomputed tomography are agglomerations of smaller calcified matrix vesicles. µCalcs < 5 µm are predicted to be not harmful, because the tiny voids associated with these very small particles will not explosively grow under tensile forces because of their large surface energy. These observations strongly support the hypothesis that nearly all fibrous caps have µCalcs, but only a small subset has the potential for rupture.