Individual Red Blood Cell Fetal Hemoglobin Concentrations Determined By Imaging Flow Cytometry and the Number of F Cells Are Very Good Predictors of Clinical Response to Increased Level of HbF and Are Associated with Hemolysis, Tricuspid Regurgitant Jet Velocity, and Sickle Retinopathy in Adult Sickle Cell Disease Patients

Abstract

Sickle cell disease (SCD) patients with similar total fetal hemoglobin (HbF) levels may present with different severity, which is attributed to the heterogeneous distribution of HbF among red blood cells (RBCs) with varying percentages of circulating HbF-containing RBCs (% F cells). Steinberg et al. (2014) proposed that an individual HbF concentration per cell (iHbF) above 10pg inhibits the polymerization of HbS, so the development of precise methods to quantify iHbF and the percentage of F cells may allow clinicians to use those parameters as targets to monitor patient treatment. We devised a novel procedure to very accurately measure and quantify iHbF using imaging flow cytometry (Amnis Imagestream) and investigated its correlations with clinical manifestations. Samples from 54 adult SCD patients (22 [41%] female, 39 [72%] on hydroxyurea) were used to acquire images from RBCs labeled with fluorescent anti-HbF antibody and were analyzed using IDEAS software v 6.2. Along with the hematimetric data and HPLC quantification of total HbF of each sample, we modified the formulas by Maier-Redelsperger et al. (1994) and Horiuchi et al. (1995), with minimization of interference from autofluorescence of non-F cells, correction for area and fluorescence intensity, and analyzed RBC subpopulations with varying levels of HbF expression. Statistical analysis was performed using R considering a p value <0.05 to be statistically significant. We did not find associations of iHbF with alpha thalassemia carrier state or beta globin cluster haplotypes Benin or Bantu, the most common in our population. Perpendicular cutoff correlation studies defined two patient populations: one group with lower mean iHbF (9.9pg) and lower % F cells (28.2%) and another group with a higher mean iHbF (15.8pg) and higher mean % F cells (72.5%). The latter group had a lower frequency of vaso-occlusive crises (p=0.011). That was further validated with our finding that the cutoffs of % F cells>30% and iHbF>10pg were strongly associated with lower absolute reticulocyte counts (p<0.001), LDH (p=0.002), and indirect bilirubin (p<0.001). The association between hemolysis and iHbF was further explored regarding pulmonary hypertension, showing that a value of iHbF>10pg was associated with a lower median tricuspid regurgitant jet velocity than in patients with iHbF<10pg (2.72m/s vs. 2.29m/s, p=0.046). We also found an association of a mean iHbF<8pg with the presence of sickle retinopathy (p=0.002). Patients with retinopathy had a lower mean iHbF and % F cells than those without it (7.08pg vs. 8.75pg, p=0.032; 48.5% vs. 71.0%, p=0.043, respectively), suggesting increments in iHbF and % F cells may be protective against sickle retinopathy, a complication not usually associated with high hemolytic rates. Our results suggest that a percentage of F cells above 30% along with individual RBC HbF concentration of 10pg may be protective against complications in SCD patients and, therefore, should be explored as goals for hydroxyurea therapy in this population. In addition, the data presented here suggest further studies on the efficacy of HbF induction therapy against the development of sickle retinopathy may be performed in a large population of sickle cell disease patients.