(A-322) Associations between lisinopril and amlodipine use and risk of aneurysmal subarachnoid haemorrhage: A UK population-based cohort study

Background: The aneurysmal subarachnoid haemorrhage (aSAH) is often a fatal type of stroke, with only high-risk surgical procedures available, and an urgent need for non-invasive drugs preventing it. A previous drug-wide association study showed that lisinopril and amlodipine might be associated with a reduced risk of aSAH (Kanning et al 2024).

Objectives: We aimed to investigate the association between lisinopril, and amlodipine use and the incidence of aSAH in a large real-world data study from the UK.

Methods: This was a retrospective cohort study, including adult patients (≥18 years old) with hypertension from the Clinical Practice Research Datalink GOLD database linked to Hospital Episodes Statistics and Lower Layer Super Output Areas data between 01.08.2004 and 31.03.2021. Adopting an active-comparator new-user design, we included the initiators of lisinopril and other angiotensin converting enzyme inhibitors (ACEis) in the 1st cohort, and initiators of amlodipine and other dihydropyridine calcium channel blocker (dCCBs) in the 2nd cohort, and followed them up from the first prescription date (index date) until the end of treatment or switch to the comparator (following a per-protocol design), aSAH diagnosis, death, disenrollment from database, or end of study period, whichever came first. Standardised mean differences of lifestyle factors, comorbidities, and socioeconomic status were calculated, and propensity scores (PS) were estimated, with PS distributions visualised, to test the comparability of the two exposure groups in each cohort. Multivariable Cox regression models were fitted to estimate the risk of aSAH, controlling for actual confounders from the PS analysis.

Results: We identified 40,187 lisinopril (mean age 60 years, 51% men), and 116,054 other ACEi users in the 1st cohort, and 116,347 amlodipine (mean age 63 years, 51% men) and 27,920 other dCCB users in the 2nd cohort. During follow-up, 14 aSAH cases occurred among lisinopril users (incidence rate [IR] 17 per 100,000 person years [PYs]), and 63 among other ACEis (IR 27 per 100,000 PYs). There was a 39% decreased risk of aSAH among lisinopril users compared with other ACEis, but not statistically significant (hazard ratio [HR] 0.61, 95% CI 0.35-1.10). We also found 31 aSAH cases among amlodipine users (IR 15 per 100,000 PYs) and 12 among other dCCBs (IR 21 per 100,000 PYs), with a non-significant 31% decreased risk of aSAH with amlodipine use versus other dCCBs (HR 0.69, 95% CI 0.36-1.36).

Conclusions: According to the per-protocol analysis, both lisinopril and amlodipine might be associated with a decreased risk of aSAH compared to other drugs from the same class, but our risk estimates lacked enough precision. The results of the on-treatment design, and dose investigations will complement this.