Risk Profile of Symptomatic Lacunar Stroke Versus Nonlobar Intracerebral Hemorrhage
Background and Purpose—Although lacunar stroke (LS) and deep intracerebral hemorrhage (dICH) represent acute manifestations of the same pathological process involving cerebral small vessels (small vessel disease), it remains unclear what factors predispose to one phenotype rather than the other at individual level.
Methods—Consecutive patients with either acute symptomatic LS or dICH were prospectively enrolled as part of a multicenter Italian study. We compared the risk factor profile of the 2 subgroups using multivariable logistic regression.
Results—During a time course of 9.5 years, 1931 subjects (1434 LS and 497 dICH; mean age, 71.3±13.3 years; males, 55.5%) qualified for the analysis. Current smoking was associated with LS (odds ratio [OR], 2.17; P<0.001). Conversely, dICH cases were more likely to be hypertensive (OR, 1.87; P<0.001), excessive alcohol consumers (OR, 1.70; P=0.001), and more frequently under treatment with warfarin (OR, 2.05; P=0.010) and statins (OR, 3.10; P<0.001). Hypercholesterolemia, diabetes mellitus, and antiplatelet treatment were not associated with a specific small vessel disease manifestation.
Conclusions—The risk factor profile of dICH differs from that associated with LS. This might be used for disease risk stratification at individual level.
The term cerebral small vessel disease (SVD) encompasses a wide spectrum of pathological processes affecting the small vessels of the brain,1 among which, arteriolosclerosis, also known as hypertensive microangiopathy, is the most common. Lacunar strokes (LS) and hemorrhages located in the deep brain regions (nonlobar or deep intracerebral hemorrhage [dICH]) represent the acute symptomatic consequences of this progressive microangiopathy. Because these two phenotypes share the same pathological substrate,1 it is still matter of debate why some individuals are more prone to develop acute small vessel occlusion, whereas others are more susceptible to vascular rupture. To date, few studies have attempted to address this issue, with conflicting results. Most of the inconsistency arise from methodological drawbacks of the study designs, including the relatively small number of subjects in some analyses, the exclusion of subjects without hypertension in others, as well as the lack of a clear distinction between ischemic stroke subtypes and between lobar and dICH.2–5 Therefore, in this study we aimed at investigating whether differences in individual risk factor profile underlie the propensity for developing dICH, rather than LS.
For the purpose of the present analysis, we reviewed data from the prospectively collected registries of 4 Italian neurological centers. Consecutive patients who were admitted between January 2002 and July 2011 were considered if their presenting symptoms were consistent with a first-ever acute cerebrovascular event. Eligibility for study participation required neuroimaging (computed tomography or magnetic resonance imaging) confirmation of LS or dICH. Patients’ selection criteria and risk factors definition are provided in the online-only Data Supplement.
Categorical variables are reported as proportion and continuous variables as mean (SD). Descriptive differences between groups were examined with the χ2 test and 1-way ANOVA F test, when appropriate. Multivariable logistic regression was performed including demographics (age and sex), as well as hypertension, diabetes mellitus, hypercholesterolemia, current smoking, antiplatelet therapy, use of oral anticoagulants, statin treatment, and alcohol consumption as covariates. P<0.05 on 2-sided test was considered significant. Data were analyzed using SPSS v.16 (http://www.spss.com).
A total of 1931 patients fulfilled the inclusion criteria (mean age, 71.3±13.3 years; males, 55.5%). Of these, 1434 (74.2%) had LS and 497 (25.8%) had dICH. The characteristics of the whole study group are summarized in Table 1.
Patients presenting with dICH were more likely to be hypertensive, to have a history of ischemic heart disease, excessive alcohol consumption, as well as statin and oral anticoagulant treatment compared with LS subjects. However, current smoking and hypercholesterolemia were significantly more frequent in the group of patients with LS.
Hypertension, excessive alcohol use, anticoagulant, and statin therapy were associated with dICH in the multivariable logistic regression model, whereas current smoking was associated with LS (Table 2).
The main finding of this study is that dICH and LS exhibit distinct risk factor profiles.
First, our study corroborates previous findings indicating hypertension as major risk factor for dICH.2,6 Our results are indirectly supported by previous reports suggesting that the benefit of long-term antihypertensive therapy might be greater for the prevention of hemorrhagic stroke than of brain ischemia.7 Second, as opposed to hypertension, diabetes mellitus was not associated with a specific clinical manifestation of SVD in our study. This is in an apparent disagreement with previous reports.4 Such discrepancy might be explained by the lack of a precise identification of ischemic stroke subtypes and of pathogenic subgroups of cerebral bleeding based on hemorrhage location (deep versus lobar). Third, consistently with previous evidence that excessive alcohol intake is an important susceptibility factor for cerebral bleeding,8 excessive alcohol consumption turned out to be related to dICH in our analysis. Conversely, in line with previous findings from Kaplan et al,9 current smoking was associated with the occurrence of LS. Fourth, as expected, we observed a strong relationship between oral anticoagulant therapy and dICH,6 as opposed to a neutral effect of antiplatelet medications. Finally, statin use seems related to the hemorrhagic manifestation of SVD. Although the benefits of statin treatment in primary and secondary prevention of ischemic stroke are widely recognized, the association between these molecules and the risk of cerebral bleeding is still controversial.10 Although this study was not specifically designed to investigate this hypothesis, our results are in agreement with previous reports suggesting a possible relationship between statin therapy and increased risk of ICH.10 Statins pleiotropic effects, including reduction of platelet aggregation and inhibition of the coagulation cascade,11 make the relation between statins and cerebral bleeding in subjects with underlying SVD biologically plausible.
The major strengths of our study are the large number of participants, the homogeneous demographic characteristics of the cohort along with the standardized diagnostic workup and evaluation of risk factors. Some limits should also be acknowledged. The lack of a systematic magnetic resonance imaging screening prevented us from investigating the potential role of imaging markers, such as white matter lesions and cerebral microbleeds. Hospital referral selection bias and the long inclusion period may also have affected our analysis. Finally, because of the lack of a stroke-free control group, we cannot but speculate that the observed associations might represent causality.
Symptomatic clinical manifestations of cerebral SVD may depend on distinct risk factor profiles predisposing to bleeding and thrombosis, respectively.
Dr Zini received speaker fees from Boehringer-Ingelheim, Medtronic-Covidien and consulting fees from Nestec, Medronic-Covidien.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.013722/-/DC1.
- Received April 8, 2016.
- Revision received May 7, 2016.
- Accepted May 18, 2016.
- © 2016 American Heart Association, Inc.
- O’Donnell MJ,
- Xavier D,
- Liu L,
- Zhang H,
- Chin SL,
- Rao-Melacini P,
- et al
- Andersen KK,
- Olsen TS,
- Dehlendorff C,
- Kammersgaard LP
- Marsh EB,
- Gottesman RF,
- Hillis AE,
- Maygers J,
- Lawrence E,
- Llinas RH
- Bonetti PO,
- Lerman LO,
- Napoli C,
- Lerman A