Therapy Insight: Stroke Risk and Its Management in Patients With Sickle Cell Disease

Fenella J. Kirkham


Nat Clin Pract Neurol. 2007;3(5):264-278. 

In This Article

Risk Factors for Stroke in Sickle Cell Disease

There seems to be a familial predisposition to stroke[55] and to high blood flow velocities[56] in SCD, indicating that genetic factors probably contribute to stroke risk. Siblings might, however, also share adverse environmental conditions, including poverty, air pollution and poor nutrition. There is considerable current interest in epistatic polymorphisms as additional risk factors for stroke in SCD. The A3 and A4 alleles of the GT-repeat polymorphism of the angiotensinogen gene seem to be associated with a fourfold increase in risk of clinical stroke in SCD, perhaps because of an effect on blood pressure.[57] Factor V Leiden and the thermolabile variant of the methylenetetrahydrofolate reductase gene are rare in Afro-Caribbean populations, and at present there is little evidence for a link between stroke and these or any other genetically determined or acquired prothrombotic disorders in patients with SCD.[57]

In recent years, evidence for human leukocyte antigen (HLA)-linked susceptibility for stroke has been found for the class I HLA-B and class II HLA-DRB loci in children with SCD. Hoppe and her co-workers[58] used a candidate gene approach in patients screened with MRI as part of the US-based CSSCD. Dividing patients on the basis of the MRI findings into three groups—those with normal imaging, those with infarction clearly involving the territories of the large vessels (with or without additional small-vessel involvement), and those with infarction not involving the territories of the large vessels (presumed small-vessel disease)—the authors found that polymorphisms in specific HLA types and in genes involved in inflammation, cellular adhesion, blood pressure regulation and lipid metabolism were differentially associated with the two stroke phenotypes. Another group has demonstrated that large-vessel disease in SCD is related to variations in genes that code for factors involved in the responses to inflammation, hypoxia, adhesion and coagulation.[59]

Sickle erythrocytes adhere more avidly to vascular endothelial cells than do normal erythrocytes, and in SCD evidence has also accrued for neutrophil and platelet adhesion to the vascular endothelium, and activation of pathways leading to inflammation and thrombosis. Vascular cell adhesion molecule 1 (V-CAM1) is of particular interest in SCD as this molecule seems to coordinate the inflammatory response by recruiting leukocytes, and mice depleted of V-CAM1 have a high leukocyte count—a risk factor for stroke in humans.[27] Interestingly, there is a high frequency of single nucleotide polymorphisms (SNPs) in the VCAM1 gene in people of Afro-Caribbean origin, and one of these variant alleles (Gly1238Cys) has been shown to associate with protection from stroke in the Jamaican population, as compared with the wild-type allele.[60]

Sebastiani et al. have recently used Bayesian modeling to examine the association of SNPs in candidate genes with sickle-cell-related stroke.[61] They found 31 SNPs in 11 genes, including bone morphogenetic protein 6 (BMP6), three genes involved in the transforming growth factor β (TGFβ) signaling pathway, and SELP (P-selectin; also known as granule membrane protein 140 and antigen CD62), which is known to be associated with stroke in the general population; all of these factors seemed to directly affect stroke risk. They also identified SNPs in a further nine genes, including endothelin-1, which is close to BMP6 on chromosome 6 and is upregulated during acute hypoxia, that seemed to be acting indirectly. When validated in a separate population, this combination of genes, interacting with the percentage of hemoglobin F, was found to have 98.2% accuracy for distinguishing patients with stroke from those without stroke, with all seven of the studied strokes correctly classified.

Low hemoglobin, high white cell count, previous transient ischemic event, hypertension and history of chest crisis all seem to be risk factors for overt ischemic stroke in SCD ( Table 2 ).[3,27] For hemorrhagic stroke, only low hemoglobin and high white cell count were found to be predictors in the CSSCD ( Table 2 ).[3] A recent study comparing hemorrhagic with ischemic stroke in children, however, did not confirm these two variables as risk factors, and instead highlighted the importance of recent transfusion and corticosteroids, perhaps in relation to acutely increased blood pressure.[20] Large-vessel disease might be associated with markers of increased hemolysis, such as lactate dehydrogenase[62] and reticulocyte count.[63] Patients presenting with neurological symptoms after chest crisis are more likely to have an atypical stroke syndrome or stroke mimic—such as hemorrhagic stroke,[14] posterior circulation stroke,[14,63,64] symptomatic or asymptomatic (covert) single or multiple small infarcts or 'lacunae',[14] reversible posterior leukoencephalopathy,[14] global or focal edema,[63] laminar necrosis[14] or acute necrotizing encephalomyelitis[15]—than they are to have a typical infarctive stroke; thrombocytopenia was the only predictor for neurological symptoms in a large series of patients with ACS.[13] Covert infarction is also associated with high pocked red cell count in infancy, compatible with early splenic infarction, as well as low hemoglobin and high white cell count; interestingly, covert infarction is less common in individuals with frequent pain, but more common in those with a past history of seizures[32] ( Table 2 ). Progressive covert infarction is more common in patients with a persistently high white cell count,[54] suggesting a role for chronic inflammation. Factors associating with low IQ and other cognitive problems in SCD include hematological variables such as low hemoglobin,[29,65,66] high white cell count[66,67] and thrombocytosis,[29] in addition to parenchymal brain damage,[32,34,37,39] large-vessel disease,[63,68] perfusion abnormality[67] (Figure 3B) and nutritional factors (see below).

In African Americans, there is a high prevalence of individuals with only two or three α globin genes (compared with the normal complement of four). This α thalassemia seems to reduce the risk of stroke, probably through an increase in hemoglobin levels.[69] High hemoglobin F levels seem to ameliorate the risk of overt stroke and silent infarction,[70] at least in childhood. The β globin haplotypes might also alter risk, probably by influencing hemoglobin F levels, although the data are conflicting. Increasing hemoglobin F levels provides the rationale for the use of hydroxyurea to ameliorate disease severity ( Table 2 ).

Recent data indicate that nocturnal hypoxemia, related in part to anemia, might increase the risk of CNS events in patients with SCD.[12] Platelet and leukocyte activation, which might affect endothelial function, are inversely related to mean overnight oxygen saturation in children with SCD,[71] and there is evidence for increased levels of markers of cellular and endothelial adhesion.[72] The mean overnight oxygen saturation seems to be associated with the severity of intracranial vasculopathy determined by turbulent flow on MRA,[63,73] which is in turn related to regional cerebral perfusion.[9] Brain areas with low perfusion are at risk of ischemia, which might be exacerbated if there is also hypoxemia. It is not yet clear if this effect is secondary to sustained hypoxemia or obstructive sleep apnea, which have been documented in reports of children with arterial ischemic stroke. Although the numbers of patients studied to date have been small, adenotonsillectomy—commonly used to treat obstructive sleep apnea or recurrent tonsillitis—might not reduce the risk of CNS events,[12] perhaps because hypoxemia commonly persists after surgery. There is some evidence that asthma predisposes to chest crisis and CVA,[74] and the complex relationships between these comorbidities are currently under intense investigation. A pilot feasibility and safety trial of overnight respiratory support has commenced.[75]

Acute and chronic infections have long been recognized as precipitants for the neurological complications of SCD; however, there are few data on the effects of penicillin prophylaxis or vaccination against Streptococcus pneumoniae or Haemophilus influenzae, which are now the standard of care. For the neonatally screened East London SCD cohort, the introduction of penicillin prophylaxis from infancy was associated with 95.2% stroke-free survival at 15 years,[76] which seems to be an improvement compared with previous reports.[3,77] Aspirin might prevent stroke, silent infarction and cognitive impairment by mechanisms that include reduction of inflammation and the antiplatelet effect. A pilot trial has commenced;[75] however, until a phase III randomized trial has been completed, aspirin must be used with caution in patients with existing ischemic stroke and cerebrovascular disease because of the risk of hemorrhage. Blood pressure is relatively low in patients with SCD in comparison with controls, but relative hypertension is associated with increased risk of ischemic stroke.[3] The effect of controlling blood pressure on stroke risk has not, however, been examined systematically.

There are few data on the role of nutrition in determining the risk of overt and silent stroke in SCD, although there is increasing evidence for an effect in elderly adults. If nutritional factors do indeed contribute to stroke risk in patients with SCD, this might explain certain geographical variations that have been observed. For example, although Greek patients with HbSβ0-thalassemia have silent infarction, high TCD velocities and cognitive problems seem to be rare;[78] this rarity might be related to the Mediterranean diet. Short stature was a risk factor for low IQ in the Jamaican cohort,[79] but these patients did not undergo MRI, so the existence of any interaction with covert infarction could not be determined. In the East London cohort, low IQ was associated with infarction on MRI and with cerebrovascular disease and, in addition, was independently predicted by short stature from the age of 3 years.[80] The data on the relationships between anthropometric measurements and the risks of large-vessel disease or overt stroke are limited.

There is considerable evidence that high levels of homocysteine, related to genetic factors as well as vitamin intake, predispose to cerebrovascular disease, including accelerated atherosclerosis, dissection, thrombosis, embolism and venous sinus thrombosis. Some studies have shown high homocysteine levels in children with SCD,[81,82] which seem to be correctable with vitamin B supplementation,[82] although there is controversy over the relevance of these findings to vascular disease. Homocysteine levels might be reduced by folate, vitamin B12, pyridoxine (vitamin B6) and riboflavin (vitamin B2) supplementation. The importance of diagnosing pernicious anemia, particularly in patients on folate supplementation, was underscored by a recent case report.[83] Iron deficiency is a common comorbidity of SCD in young children in the developing world.[84] This deficiency is likely to have a complex effect on phenotype, reducing the degree of hemolysis and potentially its associated complications,[62] but possibly having a detrimental effect on cognitive function in infancy.[36]

Other supplements might improve endothelial function in patients with SCD. Morris et al. have shown that pulmonary pressures were reduced by arginine supplementation in patients with pulmonary hypertension;[85] however, there are currently no data on cerebrovascular disease or neurological complications. Intake of antioxidants such as aged garlic, ascorbic acid and vitamin E might affect intermediate risk factors, such as oxidative damage,[86,87] hematology,[87,88]blood pressure[87] and the hypoxic hyperventilatory response;[63,89] however, there are no data on the effects of supplementation on endothelial function, cerebrovascular disease or the risk of neurological events. Although increased hemolysis is a theoretical concern with high-dose vitamin C,[90] there is no evidence that it is an actual side effect.[91] Zinc supplementation, which reduces red blood cell dehydration, might be associated with reduced hemolysis and other crises,[92] but there are no data on its effects on neurological disease. There are very few data on other nutritional supplements, such as fish oil,[91] which might be of benefit to the general population at risk of vascular disease. Trials of appropriate nutritional supplementation from infancy would be justified as there is evidence for low levels of key components in individuals with SCD despite adequate dietary intake as in the general population, indicating that consumption needs to be increased in this group of individuals. Currently, as there are no known nutritional risk factors, health professionals caring for patients with SCD should encourage consumption of a wide variety of foodstuffs as part of a healthy diet, including at least five portions of fruit or vegetables per day.

Recent data from the CSSCD showed that covert infarction seen on MRI was associated with an increased risk of overt stroke (1.03 per 100 patient-years) and progression of covert infarction (7.06 per 100 patient-years).[28] The Silent Infarct Transfusion Trial, in which children with covert infarction seen on MRI will be randomized to blood transfusion or observation, is currently enrolling patients and will report after 2011.[75]

TCD is a safe, noninvasive, well tolerated, relatively low-cost procedure in which the velocity of blood flow can be measured in intracranial vessels using an ultrasound probe placed over the temporal bone to screen for cerebrovascular disease.[53] In a comparison with conventional angiography, TCD showed a sensitivity of 90% and specificity of 100% for the diagnosis of abnormality.[93] Follow-up studies have indicated that blood velocities over 200 cm/s (abnormal) and between 170 cm/s and 200 cm/s (conditional) in the internal carotid or middle cerebral artery of children with SCD predict stroke risks of 40% and 7%, respectively, over the subsequent 3 years,[53] presumably because high velocities indicate the development of severe middle cerebral artery narrowing secondary to turbulence or fixed stenosis. TCD might become abnormal before MRA does;[94] however, although it is more expensive and can require general anesthesia in young children, MRA is very useful in confirming the presence and extent of cerebrovascular disease.[10,50,51,52,94]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.