Abstract Introduction Sickle cell disease (SCD) is characterized by chronic hemolysis and endothelial dysfunction, driving inflammation and oxidative stress. Red blood cell (RBC)-derived extracellular vesicles (RBC-EVs) and plasma-derived EVs (pEVs) contribute to endothelial activation and abnormal RBC adhesion, yet their role in SCD pathophysiology and treatment response remains underexplored. This study investigates whether EV profiles reflect disease severity and treatment-specific responses, potentially serving as non-invasive biomarkers for inflammation, hemolysis, and therapeutic outcomes in SCD patients treated with hydroxyurea (HU), voxelotor, or transfusion. Materials and Methods We isolated RBC-EVs and pEVs from SCD patients receiving HU (n=4), voxelotor (n=5), or transfusion (n=11), and healthy controls (n=3). EVs were extracted using polymer-based precipitation, with concentration and size measured via nanoparticle tracking analysis (NTA) and structure and morphology verified by cryo-electron microscopy (Cryo-EM). Clinical biomarkers were used to calculate BioScore(lactate dehydrogenase, C-reactive protein, ferritin) and BloodScore(RBC, white blood cells, platelets, mean corpuscular volume, hematocrit). Spearman's correlation assessed relationships between EV characteristics and clinical parameters, including fetal hemoglobin (HbF). Results and Discussion pEV concentrations were significantly lower in SCD patients (1.29 x 1012 particles/mL) compared to controls (2.18 x 1012 particles/mL, p=0.0006), but SCD pEVs were larger (87.2 ± 13 nm vs. 69.1 ± 9 nm, p0.0001). Transfusion and voxelotor groups had pEVs levels closer to controls, while HU-treated patients showed markedly reduced pEV concentrations (4.66 x 1011 particles/mL). No significant differences were observed in RBC-EVs concentration or size across treatments. Generally, BioScore weakly correlated with pEV (ρ=0.35) and RBC-EVs (ρ=0.39) concentrations, but presented stronger treatment-specific correlations in voxelotor (ρ=0.67) and HU (ρ=0.40) groups. Notably, rEV concentration is strongly negatively correlated with BioScore (ρ=-0.87), suggesting inflammation suppresses RBC-EVs release. pEV size was negatively associated with BioScore in HU (ρ=-0.56) but positively in voxelotor (ρ=0.56) and transfusion (ρ=0.42). In HU-treated patients, HbF strongly correlated with pEV (ρ=0.80) and rEV (ρ=0.87) concentrations and negatively with pEV size (ρ=-0.80), while voxelotor showed moderate negative correlations (pEV: ρ=-0.56; size: ρ=-0.67). No significant HbF correlations were found in transfusion patients. These findings indicate EVs capture distinct treatment-specific responses, reflecting inflammation and hematological remodeling. Conclusion Our findings underscore the innovative potential of EVs as dynamic, non-invasive biomarkers that bridge the gap between molecular mechanisms and clinical management in sickle cell disease. We observed distinct associations between treatment modalities, clinical parameters, and EV dynamics, particularly in response to hydroxyurea. Unlike conventional laboratory markers that offer a limited and static view of patient status, EVs capture real-time biological activity within the vasculature and cellular environment. This capacity enables earlier identification of disease exacerbation, more precise monitoring of therapeutic response, and supports real-time clinical decision-making. Moreover, the feasibility of incorporating EV analysis into point-of-care platforms positions them as promising tools for advancing precision medicine and improving patient outcomes in SCD.
Tincher et al. (2025) studied this question.