Candidate Gene Associations with Vaso-Occlusive Crisis in Children with Sickle Cell Anemia Enrolled in the Multinational DOVE Trial

Introduction: Sickle cell disease (SCD) is complex, with marked inter-individual variability in disease severity. The clinical heterogeneity of SCD is largely influenced by genetic factors, including several well-established modifiers of fetal hemoglobin (HbF) levels within the beta globin locus, BCL11A, and HBS1L-MYB genes. However, little is known about genetic factors that may influence variability in treatment response, particularly in clinical trials assessing potential therapeutic agents for SCD. The DOVE (Determining Effects of Platelet Inhibition on Vaso-Occlusive Events) study was a multinational, phase 3, randomized, double-blinded, placebo-controlled trial evaluating the efficacy and safety of the antiplatelet agent prasugrel for the reduction of vaso-occlusive crisis (VOC) in children with sickle cell anemia (NCT01794000). Although the primary endpoint of reduction of rate of VOC did not reach statistical significance in the DOVE trial in the ITT population, a treatment effect of prasugrel was suggested in certain subgroups. The objective of this study was to evaluate candidate genetic variants with SCD severity and to assess impact of these variants on response to prasugrel, as measured by the frequency of VOC (composite of painful crisis and acute chest syndrome events) during the study period of up to 24 months.

Methods: De-identified dried blood spot samples were collected for genetic analyses as per informed consent and study protocol. Genomic DNA was extracted from a 3 mm Whatman FTA card punch using Qiagen QIAamp DNA kit. A total of 313 (92%) DOVE patients (Hb SS or Hb S/β⁰ thalassemia) were genotyped using the TaqMan assay. After quality control, 253 patients and 28 variants were considered for genetic evaluation. These variants have been reported to be associated with SCD and/or its related comorbidities. All variants were analyzed for VOC in the prasugrel and placebo groups separately and for treatment by variant interactions in the overall study population. Variants were also assessed in the overall study population for associations with a hemolytic score, derived from 5 biomarkers of hemolysis and baseline hematologic markers including leukocyte, neutrophil, and platelet counts, and red cell mean corpuscular volume. All analyses were conducted using additive multivariate regression framework as the main model and were adjusted for age, race, geographic region, and hydroxyurea (HU) use at baseline. In addition, genotypic, dominant, and recessive models were evaluated as supportive analyses.

Results: Two variants, rs7482144, located in the 5'- HBG2 gene promoter (-158 Xmn1) on chr 11 and rs1427407, located in the BCL11A gene on chr 2, were associated with VOC and a treatment interaction effect of prasugrel. HBG2 rs7482144, was associated with a significant decrease in frequency of VOC in prasugrel-treated patients relative to placebo (adjusted p=0.0065). The BCL11A rs1427407 variant was associated with an increase in VOC in prasugrel-treated patients relative to placebo (adjusted p=0.0205).

No genetic associations were found with baseline composite hemolytic score in the overall population. However, significant associations with KLF-1, HBG2, and HBE variants were observed with individual hematologic markers. The KLF-1 variant (rs112631212) was associated with lower leukocyte (p<0.0001) and neutrophil (p=0.0002) counts in the subgroup of participants on HU; whereas the HBG2 (rs2070972; p=0.0116) and HBE (rs3834466 and rs7130110; p=0.0274 and p=0.0340 respectively) variants were associated with elevated platelet counts in the overall group.

Conclusions: Given the study sample size, this study highlights the significance of HBG2 and BCL11A variants for the frequency of VOC in prasugrel-treated patients versus placebo. These variants have been reported in association with HbF/F-cells, an ameliorating factor for VOC that is only partly explained by endogenous levels of HbF. Our findings suggest that these variants may have contributed to the treatment-related effects of prasugrel on VOC outcomes in the DOVE trial. Identifying such genetic variants that are associated with differential response to treatment may help guide selection of patients for clinical trials based on predicted risk of VOC.