Cardiovascular and neuropsychiatric outcomes with PCSK9 inhibitors in atherosclerotic cardiovascular disease: A retrospective cohort study

Scritto il 07/07/2026
da Tianshu Gu

Br J Clin Pharmacol. 2026 Jul 7. doi: 10.1002/bcp.70687. Online ahead of print.

ABSTRACT

AIMS: PCSK9 inhibitors reduce major cardiovascular events in randomized trials of atherosclerotic cardiovascular disease (ASCVD), but real-world neuropsychiatric safety remains uncertain.

METHODS: In a propensity score-matched retrospective cohort study using the TriNetX U.S. Federated Research Network, adults with established ASCVD initiating alirocumab or evolocumab between 2017 and 2024 were matched 1:1 to statin-monotherapy initiators. The 12-month primary analysis assessed major adverse cardiovascular events (MACE: myocardial infarction, heart failure [HF], ischaemic stroke, all-cause death), neurodegenerative disease (dementia, Parkinson's disease) and psychiatric disorders (psychosis, bipolar disorder, depression, anxiety), with Bonferroni correction and negative control. Secondary analyses examined individual agents, ASCVD subtypes, statin use, an intensified lipid-lowering comparator and extended follow-up.

RESULTS: Among 60 885 matched pairs (mean age 70 years; 61% male), PCSK9 inhibitor therapy was associated with lower 12-month risks of MACE (HR 0.84, 95% CI 0.80-0.88), myocardial infarction (HR 0.82, 0.79-0.86), HF (HR 0.87, 0.83-0.92) and all-cause mortality (HR 0.40, 0.37-0.43); ischaemic stroke did not differ. After Bonferroni correction, no neurodegenerative association remained significant. PCSK9 inhibitor use was associated with higher risks of any psychiatric diagnosis (HR 1.33, 1.25-1.42), depression (HR 1.44, 1.28-1.61) and anxiety (HR 1.39, 1.28-1.50), persisting in secondary analyses; negative control outcomes did not differ.

CONCLUSIONS: PCSK9 inhibitor use was associated with favourable cardiovascular outcomes and no clear neurodegenerative harm, but higher rates of depression and anxiety. Given the observational design and potential residual confounding, statin-intolerance-related treatment selection and surveillance bias, these findings should be interpreted as hypothesis-generating and warrant further prospective evaluation.

PMID:42414099 | DOI:10.1002/bcp.70687