Abstract 4141585: Lipoprotein(a) and risk of cardiovascular disease events: an analysis in a large US national database

Introduction/Background: Despite increasing awareness of lipoprotein(a) [Lp(a)] as an independent, genetically determined, causal risk driver of atherosclerotic cardiovascular disease (ASCVD), Lp(a) screening occurs infrequently, and nationwide, comprehensive data characterizing the risk of elevated Lp(a) are lacking. Aims: To evaluate the association of Lp(a) level with cardiovascular disease (CVD) events in individuals with and without pre-existing ASCVD using real-world data from the Family Heart Database TM . Methods: Observational, retrospective cohort study using longitudinal data in over 324 million individuals from 2012-2021. Selection criteria included individuals ≥18 years with ≥1 Lp(a) test measured in nmol/L during May 1, 2013 to December 31, 2020, and ≥1 medical claim pre- and post-index date (date of earliest Lp[a] test). Lp(a) levels were categorized by percentile (<20th, 20 th to 80 th , and >80 th ). Elevated Lp(a) was defined as >80 th percentile (>140 nmol/L). Multivariable Cox Proportional Hazards model analyses compared a group with Lp(a) <20 th percentile [<10 nmol/L] and no ASCVD to five other groups that differed by pre-index ASCVD status (Yes/No) and category of Lp(a) level. Results: A total of 392,835 individuals were included: mean (SD) age 59 (13.6) years, 55% female, and 43.0% with pre-index ASCVD. Elevated Lp(a) was most prevalent in women and Black individuals. Mean/median follow-up duration was 3.0/2.5 years. In individuals without pre-index ASCVD, those with Lp(a) >140 nmol/L had an adjusted 12% increased risk of CVD events (p<0.0001) versus those with Lp(a) <10 nmol/L. Compared to individuals with no pre-index ASCVD and Lp(a) <10 nmol/L, individuals with pre-index ASCVD and Lp(a) >140 nmol/L had the highest CVD event risk: adjusted HR = 3.25, 95% CI 3.11, 3.41, P<0.0001. Conclusions: In one of largest analyses of US individuals with Lp(a) levels to date, elevated Lp(a) consistently predict the increased risk of future CVD events in individuals with and without pre-index ASCVD. These findings highlight the substantial unmet need for Lp(a) screening to support risk stratification and subsequent care.