Mass-based and molar-based Lp(a) immunoassays are interchangeable, near-identically predicting coronary and aortic valve calcification, Rotterdam Study of 5,129 finds (Am Heart J 2025)

Original title: Lipoprotein(a) immunoassays and their associations with coronary artery calcification and aortic valve calcification

Am Heart J · 18 February 2025 · 7

Leening MJG, Khan CF, Zhu F, Singh SS, Kavousi M, Sijbrands EJG, de Rijke YB, Bos D

This study, led by researchers at Erasmus MC in Rotterdam, addressed concerns that mass-based Lp(a) assays might be distorted by apo(a) isoform size and give inaccurate exposure estimates, by comparing a mass-based (Randox) and a molar-based (Roche) immunoturbidimetric assay in 5,129 unselected participants from the population-based Rotterdam Study, with a cardiac CT subset assessing coronary artery calcium (CAC) and aortic valve calcification (AVC). The two assays showed near-perfect linear correlation (R2 98.8%), diverging only at very high Lp(a) concentrations. Both assays related similarly to log-transformed CAC Agatston scores (Randox standardized beta 0.1003, P = 5.6x10-8; Roche beta 0.1004, P = 5.4x10-8) and to AVC Agatston scores (Randox beta 0.1525, P = 9.2x10-16; Roche beta 0.1539, P = 4.8x10-16). The authors conclude these mass- and molar-based immunoassays are interchangeable and yield near-identical associations with coronary and valve calcification, enabling direct comparison of Lp(a) research using either method.

Original abstract

Background: Lp(a) causes atherosclerosis and degenerative aortic valve disease, but concerns have risen that mass-based assays may beaffected by isoform sizes and provide inaccurate estimates of Lp(a) exposure.

Methods: We compared contemporary immunoturbidimetric assays reporting either mass-based (Randox) or molar-based (Roche) using data from 5,129 unselected participants from the prospective population-based Rotterdam Study cohort. We studied the association of both Lp(a) measurements with the burden of coronary artery calcium (CAC) and aortic valve calcification (AVC) in a random subset of participants who underwent cardiac CT.

Results: There was a near perfect linear correlation between Lp(a) concentrations from both immunoassays (R2 98.8%) with most pronounced differences apparent only at very high Lp(a) concentrations. Lp(a) concentrations were related with natural logtransformed Agatston scores (Randox standardized linear β 0.1003, P = 5.6·10-8; Roche standardized linear β 0.1004, P = 5.4·10-8). Lp(a) concentrations were strongly but similarly related to natural log-transformed AVC Agatston scores (Randox standardized linear β 0.1525, P = 9.2·10-16; Roche standardized linear β 0.1539, P = 4.8·10-16).

Conclusion: We demonstrate that these immunoassays provide interchangeable Lp(a) measurements, and that associations with CAC and AVC were near-identical. This provides opportunities to directly compare findings from research done with either immunoassay.

Trial Registration: The Rotterdam Study has been entered in the Netherlands National Trial Register and the WHO International Clinical Trials Registry Platform under shared catalog number NTR6831.

Summary written by lp-a.org from the published abstract; figures as published. Page updated 17 August 2026. Methods.