Risk Factors for Primary Hemorrhagic Stroke in Adults with Sickle Cell Disease.

Primary hemorrhagic stroke is an uncommon but serious complication of sickle cell disease (SCD) with mortality from 20 to 65%. Proposed risk factors include previous ischemic stroke, aneurysms, low steady-state hemoglobin, high steady-state leukocyte count, acute chest syndrome, and transfusion. We performed a retrospective case-control study to evaluate risk factors for primary hemorrhagic stroke in adults (age >19 years) with SCD hospitalized at Johns Hopkins Hospital from January 1989 to June 2007. Cases had SCD and intraparenchymal (IPH), subarachnoid (SAH), or intraventricular (IVH) hemorrhage confirmed by neuroimaging or analysis of cerebrospinal fluid; traumatic subdural and epidural hemorrhages were excluded. Controls had SCD and ischemic stroke (focal neurological deficits with corresponding cerebral infarcts by neuroimaging). Both were identified by searching the hospital discharge database using ICD-9 codes for acute stroke and SCD and reviewing divisional records. We compared continuous variables by Student’s t-test and calculated odds ratios with exact methods. We identified 7 cases (mean age 31 years, range 19 – 49, 29% male) and 9 controls (mean age 37 years, range 21 – 61, 11% male). All cases had sickle cell anemia (HbSS) and 17% had a prior overt stroke; Controls had HbSS (5/9) and HbSC (4/9) and 50% had a history of overt stroke. Cases presented with impaired mental status (5/6), headache (7/7) and seizure (5/7). Controls presented with hemiparesis (7/8) and rarely seizure (1/7). Three cases had IPH involving the frontal lobe, frontal and parietal lobes, or basal ganglia. Four patients had SAH with IVH (2) and frontal IPH (1). Cerebral angiography identified aneurysms in 3 cases. One case (14%) and no controls died during the initial hospitalization. About 50% of cases (3/6) and controls (4/9) had elevated systolic blood pressure at the time of stroke. Cases had lower steady-state hemoglobin (mean ± SEM 7.4 ± 1 g/dl vs. 9.3 ± 1.1 g/dl), lower steady-state blood pressures (systolic 120 ± 7 vs. 132 ± 11 mm Hg, diastolic 72 ± 7 vs. 73 ± 5 mm Hg) and higher steady-state leukocyte counts (12,912 ± 1007/ul vs. 11,097 ± 2520/ul) than controls, but these differences were not statistically significant. Mean hemoglobin concentration at the time of stroke increased 1.3 g/dl (22%) from steady-state in cases and 0.7 g/dl (10%) in controls. Three cases had simple transfusions (1, 4, and 11 days before their primary hemorrhagic stroke) in preparation for surgery (2) and for aplastic crisis (1). No controls were transfused, but a woman with HbSS had a hemoglobin of 14.5 g/dl at the time of stroke (from excessive erythropoietin administration). In this group of adults with SCD, primary hemorrhagic stroke was associated with genotype and antecedent transfusion. Mortality was lower than that previously described and may reflect improvements in medical care or random variation within a small sample. The contribution of antecedent events and other potentially modifiable risk factors for hemorrhagic stroke in adults with SCD deserves further evaluation.