Longitudinal high-sensitivity C-reactive protein and long-term cardiovascular outcomes in optimally-treated patients with high-risk vascular disease

Background High-sensitivity C-reactive protein (hsCRP) predicts major cardiovascular events (MACE) in patients with chronic atherosclerotic cardiovascular disease. The relationship between serial longitudinal hsCRP assessment and long-term MACE has not been explored. Purpose We tested the hypothesis that longitudinal follow-up hsCRP levels (repeated measures over time) would associate with 30-month MACE (cardiovascular death, myocardial infarction, cerebrovascular accident, coronary revascularization or hospitalization for unstable angina) rates. Methods We performed a post-hoc analysis of the ACCELERATE trial of optimally-treated patients with high-risk vascular disease and evaluated data from patients with available baseline and at least one follow-up hsCRP level measured at 3, 12, and 24 months. Those with a MACE event occurring before 3 months were omitted. Using multivariable Cox proportional hazard models, we determined the association of longitudinal follow-up hsCRP levels with MACE, its components, and all-cause mortality at 30-months, adjusting for age, race, sex, region, smoking status, body mass index, diabetes, baseline hsCRP, low density lipoprotein-cholesterol, high-density lipoprotein cholesterol, systolic blood pressure, statin use, and treatment group. Results Among 8,563 patients, mean age was 64.6±9.3 years and 22.2% were women. Median baseline hsCRP was 1.49 (0.74, 3.3) mg/L. Median time-weighted average follow-up hsCRP was 1.80 (0.88, 3.9) mg/L and mean time-weighted average follow-up LDL-C was 71.7±28.0 mg/dL. At 30 months, there was a significant stepwise relationship between cumulative MACE and increasing tertiles of last prior hsCRP value before an event or censoring (KM estimates for tertiles 1–3, respectively: 8.9 vs. 12.4 vs. 15.6%, p<0.001). In multivariable analysis, higher longitudinal follow-up hsCRP levels were independently associated with MACE [HR (95% CI) per SD: 1.19 (1.10, 1.29), p<0.001], all-cause death [1.66 (1.48, 1.86), p<0.001], cardiovascular death [1.39 (1.19, 1.62), p<0.001], myocardial infarction [1.28 (1.12, 1.46), p<0.001], stroke [1.32 (1.08, 1.62), p=0.007], and coronary revascularization [1.12 (1.01, 1.24), p=0.03], but not with hospitalization for unstable angina [0.94 (0.77, 1.14), p=0.51]. Longitudinal follow-up hsCRP was significantly associated with MACE in patients with baseline hsCRP <2mg/L [1.16 (1.04, 1.30), p=0.008] and ≥2mg/L [1.25 (1.12, 1.40), p<0.001]. Conclusions In the setting of established medical therapies, longitudinal follow-up hsCRP is significantly associated with an increased risk of long-term MACE, myocardial infarction, stroke, cardiovascular mortality, coronary revascularization, and all-cause death. Longitudinal follow-up hsCRP may thus represent a novel marker of residual cardiovascular risk, supporting further work exploring the potential diagnostic and therapeutic implications of these findings. Funding Acknowledgement Type of funding sources: Foundation. Main funding source(s): Iryna Dykun was supported by the German Research Foundation