Apolipoprotein(a) phenotypes and their predictive value for coronary heart disease: identification of an operative cut-off of apolipoprotein(a) polymorphism.

Apolipoprotein(a) isoforms of low-molecular weight are associated with coronary heart disease. However, because of the high number of apolipoprotein(a) isoforms, it is difficult to assess the cardiovascular risk linked to the apolipoprotein(a) gene of a subject; indeed a cut-off of apolipoprotein(a) polymorphism has not been established. The aim of this investigation was to identify an 'operative' cut-off that discriminates apolipoprotein(a) isoforms associated with high genetic risk for coronary heart disease. Two hundred and fifty-one patients with coronary heart disease and 284 controls were recruited. Apolipoprotein(a) isoforms were detected using a high-resolution phenotyping method. Twenty-seven apolipoprotein(a) isoforms with apparent molecular weight varying from 280 to 820 kDa were identified. Several cut-offs of apolipoprotein(a) polymorphism were used in order to compare the frequencies of apolipoprotein(a) isoforms of low and high molecular weight between patients and controls: the cut-off between 640 and 655 kDa had the highest chi 2 (130.40). Even when possible differences in apolipoprotein(a) phenotypes (subjects with at least one isoform of low molecular weight and subjects with only isoforms of high molecular weight) were assessed, the same cut-off showed the highest chi 2 (122.47). Multivariate analysis showed that apolipoprotein (a) isoforms had the greatest predictive value for coronary heart disease (F value = 107.0720), when the cut-off between 640 and 655 kDa was used. The cut-off between 640 and 655 kDa appears to be the most efficient in identifying subjects at high cardiovascular risk linked to apolipoprotein(a) gene, since this cut-off discriminates apolipoprotein(a) isoforms expressing a greater risk for coronary heart disease.