Bleeding risk of cerebral cavernous malformations in patients on β-blocker medication: a cohort study

Cerebral cavernous malformations (CCMs) are frequently diagnosed vascular malformations of the brain. Although most CCMs are asymptomatic, some can be responsible for intracerebral hemorrhage or seizures. In selected cases, microsurgical resection is the preferred treatment option. Treatment with the unselective β-blocker propranolol has been presumed to stabilize and eventually lead to CCM size regression in a limited number of published case series; however, the underlying mechanism and evidence for this effect remain unclear. The aim of this study was to investigate the risk for CCM-related hemorrhage in patients on long-term β-blocker medication.

A single-center database containing data on patients harboring CCMs was retrospectively interrogated for a time period of 35 years. The database included information about hemorrhage and antihypertensive medication. Descriptive and survival analyses were performed, focusing on the risk of hemorrhage at presentation and during follow-up (first or subsequent hemorrhage) in patients on long-term β-blocker medication versus those who were not. Follow-up was censored at the first occurrence of new hemorrhage, surgery, or the last clinical review. For purposes of this analysis, the β-blocker group was divided into the following main subgroups: any β-blocker, β1-selective β-blocker, and any unselective β-blocker.

Of 542 CCMs among 408 patients, 81 (14.9%) were under treatment with any β-blocker; 65 (12%) received β1-selective β-blocker, and 16 (3%) received any unselective β-blocker. One hundred thirty-six (25.1%) CCMs presented with hemorrhage at diagnosis. None of the β-blocker groups was associated with a lower risk of hemorrhage at the time of diagnosis in a univariate descriptive analysis (any β-blocker: p = 0.64, β1-selective: p = 0.93, any unselective β-blocker: p = 0.25). Four hundred ninety-six CCMs were followed up after diagnosis and included in the survival analysis, for a total of 1800 lesion-years. Follow-up hemorrhage occurred in 36 (7.3%) CCMs. Neither univariate descriptive nor univariate Cox proportional-hazards regression analysis showed a decreased risk for follow-up hemorrhage under treatment with β-blocker medication (any β-blocker: p = 0.70, HR 1.19, 95% CI 0.49-2.90; β1-selective: p = 0.78, HR 1.15, 95% CI 0.44-3.00; any unselective β-blocker: p = 0.76, HR 1.37, 95% CI 0.19-10.08). Multivariate Cox proportional-hazards regression analysis including brainstem location, hemorrhage at diagnosis, age, and any β-blocker treatment showed no reduced risk for follow-up hemorrhage under any β-blocker treatment (p = 0.53, HR 1.36, 95% CI 0.52-3.56).

In this retrospective cohort study, β-blocker medication does not seem to be associated with a decreased risk of CCM-related hemorrhage at presentation or during follow-up.