Different Lp(a) assay methods perform similarly for predicting valve and coronary disease, but only in white participants, a MESA study of 4679 adults (Clin Chem 2017)

Background: A number of lipoprotein(a) [Lp(a)] analytical techniques are available that quantify distinct particle components, yet their clinical efficacy has not been comprehensively evaluated. This study determined whether Lp(a) mass [Lp(a)-M], Lp(a) cholesterol content [Lp(a)-C], and particle concentration [Lp(a)-P] differentially discriminated risk of calcific aortic valve disease (CAVD) or incident coronary heart disease (CHD) among 4679 participants of the Multi-Ethnic Study of Atherosclerosis (MESA).

Methods: Lp(a)-M, Lp(a)-C, and Lp(a)-P were measured in individuals without clinical evidence of CHD at baseline. Relative risk regression and Cox proportional analysis determined associations between Lp(a) and the presence of CAVD or 12-year risk of CHD, respectively. To control for the relatively high lower limits of quantification for Lp(a)-C and Lp(a)-P assays, the upper 25th and 15th percentiles were selected as analytical cutoff points.

Results: Regardless of method or analytical cutoff, high Lp(a) concentrations were significantly associated with CAVD and CHD in MESA participants following adjustment for typical cardiovascular risk factors. Stratifying by race/ethnicity rendered most associations nonsignificant after correction for multiple comparisons, but Lp(a) remained associated with CAVD in whites irrespective of method (all P < 0.0001).

Conclusions: Associations of Lp(a)-C, Lp(a)-P, and Lp(a)-M with CAVD or incident CHD were similar in this entire MESA sample using a dichotomized statistical approach. However, the high lower limits of quantification and imprecision of the Lp(a)-C and Lp(a)-P assays limited their usefulness in our analyses and would likely do so in research and clinical settings.