Optimized automated apolipoprotein A-I assays as markers for coronary artery disease

Studies are divided as to whether or not apolipoprotein A-I (apo A-I) is a better marker for coronary artery disease (CAD) than high-density lipoprotein cholesterol. We hypothesized that the detergent Tween 20, which is thought to expose antigenic sites in apo A-I, would improve automated kit apo A-I assays as a diagnostic marker for CAD.

Apolipoprotein A-I was assayed by two standard automated methods and by the same methods after serum samples and reagents had been treated with Tween 20. Serum samples were obtained from 226 consecutive male patients, age 40-70 years, presenting for angiography, except for defined exclusion characteristics. Patients were categorized into two groups on the basis of stenosis: (1) normal, all vessels <20% stenosis, n = 79, and (2) CAD, at least one vessel >70% stenosis, n = 147. Diagnostic accuracy was assessed by receiver operator characteristic stenosis curves and forward stepwise logistic regression, where adjustment was made for significant possible confounding characteristics and drugs.

The optimal concentration of Tween 20 was found to be 0.5%. Receiver operator characteristic curves showed a greater area for apo A-I with Tween (area = 0.63 to 0.64) as compared to apo A-I without Tween (area = 0.60 to 0.62). Logistic regression indicated that apo A-I with Tween was a significantly better marker than high-density lipoprotein cholesterol. Receiver operator characteristic curves indicated that the ratio of modified apo A-I to apo B gave a significant improvement in area over the ratio of high-density to low-density lipoprotein cholesterol.

Addition of Tween 20 to apo A-I assays improved diagnostic discrimination for CAD. The modified apo A-I assays were better markers than high-density lipoprotein cholesterol, and the ratio of apolipoproteins was significantly better markers than lipoprotein lipids. These findings may explain the discrepancies between studies comparing high-density lipoprotein cholesterol and apo A-I as markers for CAD. Our data suggest that a multicenter effort toward optimizing and clinically validating apo A-I test reagents may be worthwhile.