LDL-C–lowering therapies bear a small risk of haemorrhagic stroke

Medications that regulate low‐density lipoprotein‐cholesterol (LDL‐C) levels appear to contribute to a slight increase in the risk of haemorrhagic stroke events, according to a meta-analysis.

Pooled data from 37 trials (n=284,301) showed that LDL-C–lowering therapies were associated with a 16-percent excess risk of haemorrhagic stroke (risk ratio [RR], 1.16, 95 percent confidence interval [CI], 1.01–1.32; p=0.03). [J Am Heart Assoc 2024;doi:10.1161/JAHA.123.030714]

Stratified by the type of drug used, the corresponding risk estimates for haemorrhagic stroke were 1.17 (95 percent CI, 1.01–1.36) with statins (33 trials, n=216,258), 0.86 (95 percent CI, 0.43–1.74) with PCSK‐9 inhibitors (2 trials, n=46,488), and 1.14 (95 percent CI, 0.64–2.03) with ezetimibe (2 trials, n=21,555).

“Currently available data for statins provide relatively strong evidence for a harmful effect. For nonstatin therapies, such as ezetimibe or PCSK‐9 (proprotein convertase subtilisin/kexin type 9) inhibitors, the evidence is too weak to draw conclusions regarding a potential increase in risk,” the investigators said.

Nevertheless, there was no statistically significant interaction between drug class and the observed effect (p=0.71), they added.

When analysis was limited to statins, results indicated a pattern of higher risk of haemorrhagic stroke in trials that included more older people (mean age ≥65 years; RR, 1.34, 95 percent CI, 1.04–1.73; p=0.23), fewer men (RR, 1.22, 95 percent CI, 1.01–1.48; p=0.64), or more patients with diabetes (RR, 1.49, 95 percent CI, 1.02–2.18; p=0.18).

As for triglyceride-lowering therapies (11 trials, n=120,984), there was no evidence of association with haemorrhagic stroke events (RR, 1.05, 95 percent CI, 0.86–1.30).

Clinical considerations

“Our findings confirm and strengthen the existing evidence that LDL‐C–lowering therapies increase the risk of haemorrhagic stroke,” the investigators said.

Previous meta‐analyses included fewer trials. By contrast, more trials were added to the current meta-analysis, lending increased statistical power overall and for the subgroup analyses, they noted. [Lancet 2012;380:581-90; Stroke 2012;43:2149-2156; Stroke 2021;52:3142-3150]

Addressing the clinical implications of their findings, the investigators pointed out that while statin use was associated with a 17-percent increase in the relative risk of haemorrhagic stroke, its absolute risk remained very low throughout the trials.

Additionally, the estimated number needed to harm with statin use for an average treatment duration of 6.7 years was 3,333, and the number needed to treat to prevent one ischaemic event over 5 years was 49. [Cochrane Database Syst Rev 2013;2013:Cd004816]

Therefore, statin use should not be automatically ruled out based solely on the risk of haemorrhagic stroke, especially when clinically indicated, according to the investigators.

“We need more evidence to determine whether haemorrhagic stroke cluster within a particular patient subpopulation and whether the degree of disability is comparable to ischaemic strokes. Many patients or doctors still have safety concerns about using statins, but we should encourage them to balance potential low risk of haemorrhagic stroke against expected benefits,” the investigators said.

“In patients with acute haemorrhagic stroke, current guidelines state that the effects of statins on short‐term outcome (ischaemic and haemorrhagic) are uncertain, but we could reduce this uncertainty by making individual patient data from lipid-lowering therapy trials publicly available so researchers can estimate and stratify risk/benefit by age, sex, and comorbidities,” they added. [Stroke 2022;53:e282-e361; JAMA Cardiol 2020;5:233]