Abstract 2639

Introduction:

Stroke is perhaps the most catastrophic complication of sickle cell anemia (SCA), occurring in 11% of patients with SCA before 20 years of age. There is a definite need for biomarkers that could predict which children with SCA are at greatest risk for developing these irreversible cerebrovascular events. Many candidate genetic polymorphisms have been proposed to affect stroke risk but few have been validated, mainly due to the lack of additional patient cohorts. To validate the accuracy of published genetic modifiers, we genotyped polymorphisms in two large prospective cohorts.

Methods:

Pediatric patients with SCA and documented primary stroke (n=134, average age at stroke = 5.8 ± 2.8 years) were recruited through the Stroke With Transfusions Changing to Hydroxyurea (SWiTCH, NCT00122980) study. As a control non-stroke group, pediatric SCA patients (n=104, average age = 10.2 ± 3.5 years) enrolled in the Hydroxyurea Study of Long-Term Effects (HUSTLE, NCT00305175) were analyzed. All participants in the HUSTLE cohort were over 5 years old and without previous clinical stroke prior to beginning hydroxyurea treatment. We genotyped 38 single nucleotide polymorphisms (SNP's) with published associations for stroke risk, along with α-thalassemia trait, G6PD deficiency and the β-globin haplotype of each patient.

Results:

Only 5 of the 38 candidate SNPs were associated with stroke risk (Table 1). As previously reported the presence of α-thalassemia trait was also associated with stroke risk (p=0.009). In contrast, G6PD deficiency was not associated with stroke risk. The classical β-globin gene haplotypes were determined for all 238 subjects, resulting in alleles primarily representing the four classical African haplotypes including Benin (57.6%), Central African Republic (21.8%), Senegal (9.2%) and Cameroon (3.2%), as well as atypical haplotypes (8.2%). None of the classical β-globin haplotypes were associated with stroke. However, fine-mapping of the β-globin gene locus identified recombination events within the Aγ-globin gene region, which were significantly over-represented in the stroke versus non-stroke cohorts (n=26.5% vs. n=12.5%, p=0.0001). In particular, one haplotype we term BEN-Memphis has the classical Benin background haplotype but also has recombination between the promoter and intron 2 of the Aγ-globin gene. There were significantly more stroke subjects with this novel BEN-Memphis haplotype (n=9.3% vs. n=0.5%, p<0.001).

Conclusions:

Our results confirm α-thalassemia trait is significantly protective against stroke in SCA. Fine-mapping of the β-globin gene locus identified novel recombinations within the β-globin gene locus that were associated with stroke risk. These variant haplotypes may be associated with altered γ- or β-globin gene expression. Of the other previously reported polymorphisms, only 5 of 38 SNPs were significantly associated with stroke risk (Table 1). These findings highlight the dangers of accepting non-validated genetic modifiers. The ADCY9 gene is highly expressed in the brain and is critical for neuronal signaling (Hacker BM et al., Genomics 1998). The TEK gene is an endothelial cell expressed tyrosine kinase that is crucial for prevention and recovery from stroke events (Bai Y et al., Neuroscience 2009). The ANXA2 gene has been proposed to affect the hypercoaguable state of SCA (Ling Q et al., J Clin Invest 2004). Finally, mutations in TGFBR3 have been linked with cerebrovascular disease (Santiago-Sim T et al., Stroke 2009). Further investigations at these genetic regions may help define the specific mutations that confer stroke risk or protection in children with SCA.

Table 1.

Polymorphisms with validated association with stroke phenotype.

GeneGene FunctionSNP IDHUSTLE MAFSWiTCH MAFOdds Ratiop-value
ADCY9 Cyclic AMP production rs2238432 33.2% 14.6% 0.45 0.003 
ADCY9 Cyclic AMP production rs2283497 48.1% 69.8% 1.68 0.048 
ANXA2 Plasma membrane organization rs11853426 35.1% 44.4% 2.74 0.008 
TEK Angiopoietin-1 receptor rs489347 35.6% 46.3% 2.18 0.020 
TGFBR3 Multi-functional cytokine rs284875 7.7% 13.1% 2.19 0.023 
HbA2 α-globin production rs63751476 17.3% 8.9% 0.44 0.009 
GeneGene FunctionSNP IDHUSTLE MAFSWiTCH MAFOdds Ratiop-value
ADCY9 Cyclic AMP production rs2238432 33.2% 14.6% 0.45 0.003 
ADCY9 Cyclic AMP production rs2283497 48.1% 69.8% 1.68 0.048 
ANXA2 Plasma membrane organization rs11853426 35.1% 44.4% 2.74 0.008 
TEK Angiopoietin-1 receptor rs489347 35.6% 46.3% 2.18 0.020 
TGFBR3 Multi-functional cytokine rs284875 7.7% 13.1% 2.19 0.023 
HbA2 α-globin production rs63751476 17.3% 8.9% 0.44 0.009 

The minor allele frequency (MAF) is given for each SNP. Significance between the control (HUSTLE, n=104) and stroke (SWiTCH, n=134) groups was tested using the Cochran-Armitage test. The HbA2 polymorphism is the Δ3.7kb α-thalassemia single gene deletion.

Disclosure:

Off Label Use: The off-label drug use of hydroxyurea to treat clinical complications of sickle cell anemia in children will be discussed.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution