Cyclooxygenase-2 disruption predisposes to heart failure with preserved ejection fraction

Heart failure (HF) is the adverse cardiovascular event most strongly associated with cyclooxygenase (COX)-2 selective and non-selective nonsteroidal anti-inflammatory drug (NSAID) use. However, it is unknown whether NSAID exposure results in HF with reduced (HFrEF) or preserved ejection fraction (HFpEF).

In adult mice, postnatal genetic deletion or pharmacological inhibition of COX-2 did not impair cardiac function. In contrast, aged female inducible (i)Cox-2 KO mice, compared to age- and sex-matched control mice, presented diastolic dysfunction, cardiac hypertrophy, pulmonary congestion, elevated plasma N-terminal pro B-type natriuretic peptide (BNP), while ejection fraction (EF) was preserved (≥ 50%). No such phenotype was observed in aged male iCox-2 KO mice. In aged female iCox-2 KO mice, we detected a rediversion of urinary prostaglandin metabolites toward leukotriene biosynthesis and an altered expression of calcium handling proteins in the myocardium. The phospholamban to SERCA2a ratio was augmented, inferring an inhibitory impact on SERCA2a activity.

In larval zebrafish, COX-2 inhibition by celecoxib caused a modest but significant reduction in heart rate and diastolic function, while EF was preserved. Furthermore, celecoxib increased Bnp expression and the amplitude of ventricular Ca ²⁺ transients.

In the Harvard-Partners electronic medical record (EMR), diabetic patients exposed to NSAIDs selective for COX-2 inhibition were more likely to associate with an increased risk for HFpEF than HFrEF.

Collectively, these results indicate that COX-2 deletion or inhibition does not impair systolic cardiac function, but results in a HFpEF phenotype in mice, zebrafish and humans. An imbalance of calcium handling mediates the impairment of myocardial relaxation consequent to COX-2 disruption.