Abstract 984: Change in Lp-PLA2 With Risk Change on Statin Therapy in the Collaborative Atorvastatin Diabetes Study (CARDS)
Background: Although lipoprotein-associated phospholipase A2 (Lp-PLA2) is advocated as a cardiovascular disease (CVD) risk marker, a recent review drew attention to the lack of data on whether on-statin levels reflect subsequent risk. We examined whether Lp-PLA2 changed with statin treatment in CARDS, a randomised trial that demonstrated the efficacy of atorvastatin 10mg in the primary prevention of CVD in patients with type 2 diabetes and whether baseline and on treatment Lp-PLA2 predicted acute CVD events.
Methods: We used a nested case control design sampling all those incident cases of CVD (n=210) in whom a valid plasma sample was available in the CARDS trial and randomly selecting three controls per case from those who completed the trial without a CVD event, stratified by treatment allocation. Lp-PLA2 assays were completed in 168 cases and 554 controls at baseline and in 66% of these at 1 year follow up. The median follow-up duration in the trial was 3.9 years. The diaDexus enzyme immunoassay PLAC® test, used to measure Lp-PLA2 mass, had a coefficient of variation ranging from 7.9 to 8.5% (low and high QCs). Analysis was by Cox regression and logistic regression with adjustment for treatment allocation and relevant covariates.
Results: Baseline median (interquartile range) Lp-PLA2 level was 347 ng/ml (234 – 479). Being in the top quartile for Lp-PLA2 was associated (p<0.05) positively with LDL-C, and HbA1c and inversely with HDL-C. In both the placebo and atorvastatin groups, there was no correlation between baseline and 1 year in Lp-PLA2 (Spearman’s rho<0.1). At 1 year the median within-person change in Lp-PLA2 was −58 ng/mL (−221 to +111) in the placebo group and −116 (−288 to +59) in the atorvastatin group (p<0.05 for the treatment effect). Change in LDL-C at 1 year in the treatment arm did not predict change in Lp-PLA2 (p=0.3). Baseline Lp-PLA2 did not predict CVD (HR for above median=0.92; 95% CI, 0.7–1.25; p=0.6), or CHD (HR=0.9; p=0.5) or stroke (HR=1.11; p=0.6) and neither did on-treatment Lp-PLA2 (all p>0.1).
Conclusion: Although atorvastatin treament did reduce Lp-PLA2 levels we found Lp-PLA2 to have high variability through time and it was not a useful predictor of CVD events on or off statin in this diabetic patient population.