Hemoglobin F and Incidence of Thrombotic Event in Sickle Cell Disease

Introduction

Characterized by abnormally sickle-shaped erythrocytes disrupting blood flow in small vessels, Sickle Cell Disease (SCD) continues to be the most common hemoglobinopathy in the United States, affecting approximately 100,000 people (Onimoe & Rotz, 2020). A 2017 cohort study of SCD patients found that there is an increased risk and earlier incidence of venous thromboembolisms compared to their matched controls (Brunson et al., 2017). Fetal hemoglobin (HbF) is the best characterized modifier of disease severity in SCD, and historically, patients with low HbF percentages are at a higher risk of developing vaso-occlusive complications, organ damage, and early death (Agrawal et al., 2014). We hypothesize that having a higher hemoglobin F (unaffected by sickle cell mutation), would confer some protection from hemolysis and would be inversely related to the incidence of thrombosis in sickle cell disease patients.

Methods

This study uses a descriptive and cross-sectional design to analyze existing, secondary medical record data of 496 SCD patients that received care from the Center for Blood Disorders in Augusta, Georgia. Data abstraction included history of thrombotic event (TE) and HbF values closest to the TE, along with, presence of SCD clinical complications, use of blood thinning medications, specific lab values, hemoglobin electrophoresis, and history of COVID-19. For those without a history of TE, the most recent data in these categories was collected. To test the association of HbF with a categorical variable such as medical history of thrombosis, we used a two-sample t-test. To test the association of HbF with a continuous variable such as D-dimer, we computed Pearson's correlation coefficient and used Fisher's z-test. Patients with SCD who developed thrombosis were further divided into quantiles based on HbF levels [Table 1]. We collected mean age at thrombosis and mean D-dimer for these quantiles and ANOVA was used for statistical analysis. All analyses were performed using R version 4.2.3.

Results

Of the 496 charts reviewed, 131 patients had a history of TE and 27 of those patients have more than one recorded TE (total=189). Twenty patients did not have HbF data close to TE and were not included in this analysis. The mean HbF for patients without history of TE was 8.89 (SD 10.0), whereas the mean HbF for patients with history of TE was 10.1 (SD 10.3). A two-sample t-test comparing these two groups produced a p-value of 0.2126. Mean age of thrombosis in each quantile were 31.6 (SD 14.6), 28.7 (SD 10.5), 34.8 (SD 11.7), and 37.1 (SD 11.9), respectively, with ANOVA producing a p-value of 0.008059. Mean D-dimer in each quantile were 1017 (SD 1408), 1417 (SD 1415), 1404 (SD 1493), and 2921 (SD 8898), respectively, with ANOVA producing a p-value of .6645.

Conclusions

While there is no significant relationship between mean HbF levels and development of thrombosis in this cohort of 496 SCD patients, the mean age of development of thrombosis was significantly higher in higher quantiles of HbF, compared to lower quantiles, suggesting that patients with higher HbF are protected longer from thrombosis. There was no relationship between mean D-dimer level and development of thrombosis in the patients we analyzed. In SCD patients with thrombosis, D-dimer was not significantly associated with thrombosis even between different quantiles of HbF. To our knowledge, this is the first study evaluating association between HbF level, D-dimer, and development of thrombosis in SCD patients. Current therapies for SCD focus on symptom management and disease modification. While the mechanism is still not completely understood, Hydroxyurea increases the number of red blood cells (RBCs) with detectable HbF, protecting RBCs from sickle hemoglobin polymerization, thus interrupting the natural disease pathophysiology (Steinberg, 2020). Some gene therapy for sickle cell disease aims at increasing HbF in SCD patients (Abraham & Tisdale, 2021). Would such therapy be protective against increased risk of thrombosis associated with SCD, remains to be determined. One limitation of this project is the restricted sample size of SCD patients with thromboembolism, and larger data size of sickle cell patients may better elucidate a relationship between Hemoglobin F and thrombosis in SCD patients.