Epidemiology of lobar and non-lobar intracerebral haemorrhage with focus on the association with select medications and on long-term vascular outcomes

Intracerebral haemorrhage (ICH) is the second most common type of stroke, and ICH is the most severe stroke type with the highest case-fatality rates. Approximately 80% of all non-traumatic ICH are spontaneous (s-ICH), and the most common cause of s-ICH is an underlying small vessel disease (SVD) in the form of deep perforating vasculopathy or cerebral amyloid angiopathy (CAA). Deep perforating vasculopathy is associated with non-lobar s-ICH, yet it can also cause lobar s-ICH. Lobar s-ICH can sometimes be caused by sporadically CAA, a disease that causes beta-amyloid peptide to accumulate and deposit in the cortical and leptomeningeal vessels making these vessels prone to ruptures. Therefore, the location of the s-ICH can, to some extent, reflect the underlying aetiology of the s-ICH.

Whether statins are associated with an increased risk of s-ICH has been debated for decades. Similarly, the association between various antithrombotic agents and the risk of s-ICH by haematoma location is not well investigated in large studies.  

Cohort studies have suggested that lobar s-ICH could be associated with higher rates of recurrent ICH (reICH) while the risk of ischaemic events (i.e., ischaemic stroke [IS] or myocardial infarction [MI]) seems similar between lobar and nonlobar locations. However, this has not been well established in large cohorts of unselected patients. Furthermore, whether atrial fibrillation (AF) or a history of previous vascular disease affects the risk of reICH, IS, or MI after s-ICH according to the index location is unknown.

The aim of this thesis was to provide knowledge on the association between statins and antithrombotic agents and the risk of incident s-ICH by location, and to investigate differences in the long-term course after s-ICH by the index s-ICH location. To achieve these aims, a large cohort of all incident s-ICH cases in the region of southern Denmark from 2009-2018 (approximately 2.800 patients) was identified using the Danish registries and further validated via medical journals and discharge summaries. This thesis is based on the results of three studies (Study 1-3). Study 1 and Study 2 used a nested case-control design, while Study 3 used a cohort design. In Study 1 and Study 2 cases were matched to generalpopulation controls by age, sex, and the index year. All analyses in Study 1-3 were adjusted for known confounders.

Study 1 included 989 lobar and 1,175 non-lobar cases aged ≥55 years who were matched with 39,500 and 46,755 controls respectively. Current statin use was associated with a lower risk of lobar (aOR 0.83 [95% CI 0.70-0.98]) and nonlobar s-ICH (aOR 0.84 [95% CI 0.72-0.98]). A duration-response relationship was observed with a longer duration being associated with a lower risk of s-ICH by both locations. Estimates stratified by statin intensity were comparable to the main estimates for low-medium intensity therapy (lobar aOR 0.82; non-lobar aOR 0.84), while the association with high-intensity therapy was neutral (lobar aOR 1.06; non-lobar aOR 1.02).

In Study 2, 1,040 cases with lobar and 1,263 with non-lobar s-ICH aged ≥55 years were matched to 41,651 and 50,574 controls, respectively. Platelet antiaggregant use was more strongly associated with lobar s-ICH compared to non-lobar s-ICH location (aOR 2.39 [95% CI 1.98–2.89] vs. aOR 1.75 [95% CI 1.51–2.02]). VKA use was associated with higher odds of both lobar and nonlobar s-ICH. Direct-acting oral anticoagulant (DOAC) use was more strongly associated with non-lobar compared to lobar s-ICH (aOR 3.34 [95% CI 2.33– 4.79] vs. aOR 1.66 [95% CI 1.02–2.70]). Platelet antiaggregant use were more strongly associated with probable CAA related s-ICH compared to non-probable CAA s-ICH. 

Study 3 examined the risk of reICH, IS, MI, and MACEs (major adverse vascular events including vascular death) after s-ICH by the index haematoma. This cohort study included 2,289 patients aged ≥50 years. Patients with lobar s-ICH had higher rates of reICH pr 100 person-years compared to non-lobar s-ICH (3.74 [95% CI 3.01-4.66] vs 1.24 [95% CI 0.89-1.73]), but not IS (1.45 [95% CI 1.02- 2.06] vs 1.77 [95% CI, 1.34-2.34]) or MI (0.42 [95% CI 0.22-0.81] vs 0.64 [95% CI 0.40-1.01]). The rates of reICH were similar between s-ICH locations in patients with AF. However, patients with AF and non-lobar s-ICH experienced higher rates of IS compared to patients with lobar s-ICH.

In summary, statin use, duration of use, and the intensity of treatment was associated with a lower risk of s-ICH, and the association did not vary by hematoma location. Antithrombotic agents were all associated with higher risks of s-ICH, but with varying magnitudes by location and drug type. These variations may reflect different underlying pathologies, drug interactions or a combination of these factors. Finally, lobar s-ICH was associated with higher rates of reICH than non-lobar, while there were no major differences in IS and MI between the two locations.