A Longitudinal Review Comparing Absolute Neutrophil Counts of Children with Sickle Cell Disease with and without the Duffy Null Phenotype

Background

Sickle cell disease (SCD) is an inherited hemoglobinopathy commonly affecting minority populations and causing numerous morbidities. SCD pathophysiology is partially mediated by inflammation that can increase absolute neutrophil counts (ANC). Hydroxyurea (HU) is a well-studied therapy that induces fetal hemoglobin production and reduces many SCD-related complications, especially when maximally dose-escalated. HU also reduces ANC, which is a key factor in determining HU dosing. Consequently, other factors that reduce ANC may be important to consider as they could affect HU dosing and disease course.

The Duffy null phenotype (DN), observed in ≥75% of the United States SCD population, leads to increased neutrophil egress into tissues. Although non-pathogenic, DN is increasingly recognized as a potential reason for unnecessary dose-reductions for people receiving ANC-dependent treatments. While our previous study did not reveal significant differences in ANC very early in life and in the HU treatment course, because inflammation increases with age and HU is a chronic medication, understanding its effect over time is needed to determine if DN should be a consideration for SCD management. Thus, the primary objective of this single-center, retrospective study was to determine if there was a difference in ANC among children with SCD with and without DN from ages 2-9 years who were not prescribed HU. We also compared ANC and HU dosing from 2-9 years after HU initiation between those with and without DN who initiated HU after a year-of-age.

Methods

Youth in a clinical database of patients who had erythrocyte phenotyping performed at our center from 2010-2021, confirmed to have SCD in their electronic medical records (EMR), and followed for ≥2 years for the non-HU treated analysis or ≥2 years after HU initiation for HU analyses were eligible. Those with Fy^a-Fy^b- were classified as having DN, while other phenotypes were classified as Duffy non-null (DNN). Demographic, ANC, and HU prescribing data during routine hematology visits were abstracted from the EMR. ANC for those not prescribed HU were compared between 2-9 years-of-age, whereas ANC for those who initiated HU were compared between 2-9 years after initiation.

Data were summarized with descriptive statistics. Linear mixed models with random intercepts for subjects were used to model ANC and HU dose over time. P-values <0.05 were considered statistically significant.

Results

Of the 107 subjects (58% male; median age at start period 2 years (IQR: 1.4-2.8); 57% Hemoglobin (Hb) SS, 27% HbSC, 12% HbSβ⁺, 3% HbSβ⁰, 1% other SCD; and 99% Black), 84 (79%) had DN. Among those with DN, 41 (49%) initiated HU, while 12 (52%) with DNN initiated HU. Subjects were followed for a median of 4.2 years (IQR: 3.1-7.5).

Among those not prescribed HU, mean ANC did not significantly change from age 2 to 9 years-of-age in the DN (p=0.56) or DNN groups (p=0.35). While the DN group had a lower ANC across all ages compared to the DNN group, this difference was not significant (p=0.89). However, mean ANC among the DN group was significantly lower (1.5-2.3 10³/mm³) than the DNN group at 6, 7, 8, and 9 years-of-age (all p<0.05).

Among those who initiated HU, ANC significantly increased over time (p=0.046), but the estimated rate of change did not differ between the DN and DNN groups (p=0.57). There was no difference in ANC between DN and DNN after controlling for HU dose and time on HU (p=0.34). The mean prescribed HU dose did not significantly change over time in either group, however, at 4, 5, 6, 7, 8, and 9 years after initiation, those with DN were prescribed a lower mean HU dose (3.2-3.6 mg/kg/day) after controlling for time on HU (all p<0.05).

Discussion

Our findings suggest that DN does not impact ANC among youth with SCD who are not receiving HU. While ANC may increase over time among those who initiate HU, we observed that those with DN and DNN on HU may have similar ANC trajectories and that DN may have a relatively minor impact on ANC compared to the underlying inflammatory response from SCD. The lower prescribed HU dose among those with DN after four years of HU deserves exploration but may be related to sample size. Future studies are needed to confirm our findings and examine DN’s influence on ANC at times when inflammation increases (i.e., adolescence, acute complications) to determine if DN has clinical significance for youth with SCD.