Departments of Clinical Biochemistry and Medicine^†, The University of Hong Kong and Queen Mary Hospital, Hong Kong

Recent epidemiological and clinical studies, mostly in Caucasian populations, show a strong relationship between the risk of atherosclerosis and plasma lipoprotein(a) [Lp(a)] or its closely related apolipoprotein(a) [apo(a)] concentrations. The genetic size polymorphism of apo(a) is controlled by autosomal alleles at the apo(a) gene locus. The study was designed to determine the frequency distribution of plasma apo(a) concentrations and phenotypes in Chinese patients with angiographically defined coronary artery disease (CAD) and in patients with ischemic stroke as documented by CT-scan or MRI, and to evaluate whether apo(a) gene variation is associated with coronary and cerebral atherosclerosis. The study cohort consisted of 366 healthy control subjects, 36 angiographically defined CAD patients and 43 patients with ischemic stroke. A two-site immunoradiometric assay was used to measure the plasma apo(a) concentrations and a high resolution SDS-agarose gel/immunoblotting method was employed to determine the apo(a) phenotypes (protein isoforms). Apo(a) isoforms were designated according to their number of kringle IV (K-IV) repeats in the gene. The geometric mean values of plasma apo(a) were significantly higher in CAD patients (308 U/L, 95% confidence interval, C.I.=216-439) and in patients with stroke (279 U/L, 95% C.I.=193-402) than in control subjects (164 U/L, 95% C.I.=148-181, n=366; p<0.001). The prevalence of plasma apo(a) concentrations >300 U/L was 27% in control subjects but was 56% (Odds ratio, OR=3.37; 95% C.I.=1.67-6.73; p<0.001) in the CAD patients and was 61% (OR=4.13, 95% C.I.=2.14-7.92; p<0.0001) in patients with stroke. The frequency of double-band phenotypes was quite similar among these groups (CAD, 56%, stroke, 56% vs control 58%). The CAD patients, however, had a significantly higher frequency of smaller isoforms having 11-22 K-IV repeats than the control subjects (22% vs 9%; OR=2.88, 95% C.I.=1.22-6.82; p<0.05) but the frequency of isoforms having 11-22 K-IV repeats did not differ between the control group and the stroke patients (9% vs 9%). In agreement with previous reports, our results show that an elevated plasma apo(a) value is associated with increased risk of both coronary and cerebral atherosclerosis in the local Chinese population. Smaller apo(a) isoforms (with 11-22 K-IV repeats) that were generally associated with higher apo(a) concentrations were shown to be more frequent in CAD patients than in control subjects and patients with stroke. This supports the concept that alleles at the apo(a) gene locus are predictors of the risk of CAD in most populations although whether the apo(a) level or the phenotype is the more important predictor is still unresolved.