Aims
The Standard care vs. Celecoxib Outcome Trial (SCOT) found similar risk of cardiovascular events with traditional non-steroidal anti-inflammatory drugs (NSAIDs) and the cyclooxygenase-2-selective drug celecoxib. While pre-clinical work has suggested roles for vascular and renal dysfunction in NSAID cardiovascular toxicity, our understanding of these mechanisms remains incomplete. A post hoc analysis of the SCOT cohort was performed to identify clinical risk factors and circulating biomarkers of cardiovascular events in NSAID users.

Methods and results
Within SCOT (7295 NSAID users with osteoarthritis or rheumatoid arthritis), clinical risk factors associated with cardiovascular events were identified using least absolute shrinkage and selection operator regression. A nested case–control study of serum biomarkers including targeted proteomics was performed in individuals who experienced a cardiovascular event within 1 year (n = 49), matched 2:1 with controls who did not (n = 97). Risk factors significantly associated with cardiovascular events included increasing age, male sex, smoking, total cholesterol:HDL ratio ≥5, and aspirin use. Statin use was cardioprotective [odds ratio (OR) 0.68; 95% confidence interval (CI) 0.46–0.98]. There was significantly higher immunoglobulin (Ig)G anti-malondialdehyde-modified LDL (MDA-LDL), asymmetric dimethylarginine (ADMA), and lower arginine/ADMA. Targeted proteomic analysis identified serum growth differentiation factor 15 (GDF-15) as a candidate biomarker [area under the curve of 0.715 (95% CI 0.63–0.81)].

Conclusion
Growth differentiation factor 15 has been identified as a candidate biomarker and should be explored for its mechanistic contribution to NSAID cardiovascular toxicity, particularly given the remarkable providence that GDF-15 was originally described as NSAID-activated gene-1.