Fetal hemoglobin (HbF) is recognized as a major determinant of clinical disease severity in children and adults with sickle cell anemia (SCA). Patients with elevated HbF levels have a milder disease course, and many current therapeutic protocols for SCA include pharmacological induction of HbF. However, baseline and treatment HbF levels vary widely due to presumed genetic and environmental factors. Recognized globin gene modifiers of HbF include the beta globin haplotype and a potential contribution from concomitant alpha thalassemia. To characterize more fully the influence of globin gene modifiers on both baseline and treatment HbF levels, we retrospectively determined the beta globin haplotype (Benin, CAR, Senegal, Cameroon, or Arab-Indian) by selective gamma globin gene nucleotide sequencing and the alpha globin gene number (2, 3, or 4) by PCR for 67 African-American children with SCA receiving hydroxyurea therapy at stable maximal tolerated dose (MTD). The four beta globin haplotypes and frequencies identified in our cohort of children include Benin (0.61), CAR (0.17), Senegal (0.12), and Cameroon (0.10). The number of alpha globin genes and frequencies identified were 4 genes (0.72), 3 genes (0.25) and 2 genes (0.03). Baseline and MTD HbF levels were analyzed according to each variable. The average baseline HbF value for the entire cohort of children was 7.7 ± 4.4% (median 7.6%, range 1.3 – 19.3%), while the average treatment HbF value was 23.9 ± 7.2 % (median 22.9%, range 10.2 – 40.7%). All 67 children increased their HbF in response to hydroxyurea therapy (median 16.7%, range 5.0 – 28.8%). There was a modest but statistically significant correlation between the baseline and treatment HbF (r=0.66, p<.0001). The estimated effect of one unit change in baseline HbF on treatment HbF was 1.11 (95% CI of 0.78, 1.43). When baseline %HbF was analyzed according to the beta globin haplotype, the overall ANOVA had a p-value of 0.02, indicating a statistically significant influence. Further analysis confirmed associations previously identified in adults with SCA, i.e. children with at least one copy of the CAR haplotype had a lower baseline HbF (5.9% vs 8.4%, p=.05), while those with at least one copy of the Senegal haplotype had a higher baseline HbF (11.1% vs 6.7%, p<.001). When hydroxyurea MTD (treatment) HbF values were analyzed according to beta globin haplotype while adjusting for baseline HbF, however, the effect of beta globin haplotype was not statistically significant (p=.13). Analyses of HbF according to alpha globin gene number revealed no statistically significant effects on either baseline or treatment HbF values. Taken together, these data support the hypothesis that beta globin haplotype significant influences baseline HbF values for children with SCA, but has no significant effects on hydroxyurea MTD HbF values. Accordingly, children with SCA should be offered hydroxyurea based solely on clinical indications, without consideration of baseline HbF or beta globin haplotype. Even children with low baseline HbF values or the CAR beta globin haplotype can respond to hydroxyurea therapy with an elevated %HbF. Future studies designed to identify genetic modifiers of treatment HbF values should focus on sequence polymorphisms in non-globin genes that have trans-acting effects on gamma globin gene expression.

Author notes

Corresponding author