Longitudinal Evaluation of a Standardized P-Selectin Flow Adhesion Bioassay: Potential Role for the Assessment and Prediction of Vaso-Occlusive Episodes in Sickle Cell Disease

Vaso-occlusive episodes (VOEs) in sickle cell disease (SCD) are characterized by severe pain and microvascular obstruction precipitated by adhesive interactions among endothelial cells (ECs), red blood cells, white blood cells (WBCs), and platelets. P-selectin contributes to microvascular occlusion by supporting abnormal interactions between blood cells and the vascular endothelium. Inhibition of p-selectin mediated adhesion reverses acute, VOEs in SCD mice. Additionally, SCD mice, deficient in either P- or E-selectin, develop fewer vascular occlusions. These reports led to the development of anti-adhesive therapies targeting p-selectin. For example, Crizanlizumab, an anti-P-selectin antibody, reduced the annual rate of VOEs prospectively and Rivipansel, a pan-selectin inhibitor, decreased the length of active VOEs. Objective measures of VOEs are nonexistent.

We developed a standardized, flow-based adhesion assay to measure p-selectin mediated adhesion in SCD. Whole blood (Wb) and isolated WBC samples were perfused (1.0 dynes/cm², 1.67Hz) through micro-fluidic channels. Adhered cells were enumerated and an adhesion index was calculated (total # of adhered cells/mm²). Blood samples were collected from SCD subjects (n = 35) every 3 weeks for 6 months at baseline and within 24 and 48hrs of a patient-reported VOE. Baseline and VOE states were confirmed by an electronic patient reported outcomes (ePRO) device, as previously described. VOEs were separated into 2 groups: VOEs managed at-home (Home-VOE) or during ER visit and/or hospitalization (Contact-VOE).

WB (n = 288; mean=42 ± 48 cells/mm², median=31 cells/mm²) and WBC (n = 282; mean=163 ± 149 cells/mm²; median= 116 cells/mm²) adhesion to p-selectin varied in SCD subjects (n=35) at baseline. WBC adhesion to P-selectin was significantly (p=0.36) higher during VOE (n=59; 205 ± 179) when compared to baseline (n = 282; mean=163±149 cells/mm², median=116 cells/mm²). For individual subjects, the geometric mean for WBC adhesion to P-selectin was significantly (p = 0.023) higher for VOEs vs. baseline (VOE= 109, n= 23, 95% confidence interval 94-126 vs. baseline= 148, n = 35, 95% confidence interval 114-192). Additionally, WB adhesion to p-selectin correlated with VOE frequency (time to 2^ndVOE; r = -0.71, n = 10, p = 0.02 or time between 1^stand 2^ndVOEs; r = -0.68, n = 10, p = 0.032).

WB adhesion to p-selectin positively correlated with C-reactive protein (CRP; r = 0.12, p = 0.047), reticulocyte % (r = 0.15, p = 0.017), and WBC counts (r = 0.26, p < 0.0001) at baseline. WBC adhesion to p-selectin positively correlated with hematocrit (Hct; r = -0.37, p = 0.027) and reticulocyte % (r = 0.44, p = 0.0058) during VOEs managed at home; WB adhesion to p-selectin positively correlates with CRP (r = 0.44, p = 0.011) and WBC count (r = 0.63, p < 0.0001); WB adhesion to p-selectin positively correlated with WBC count (r = 0.51, p = 0.0021) during contact-VOE.

These data confirm the normal range and longitudinal variability of SCD adhesion to p-selectin at baseline and during VOEs. WB adhesion to p-selectin was inversely related to VOE frequency although there was no difference between baseline and VOE samples. These data suggest that WB adhesion to p-selectin may be a better predictor of disease severity and not SCD symptomology. In contrast, WBC adhesion to p-selectin differentiated baseline from VOE samples suggesting this assay may be a more suitable reflection of SCD clinical state. Both WB and WBC adhesion to VCAM-1 correlates with biomarkers of hemolysis and inflammation. P-selectin may be selective for the identification of pro-adhesive phenotypes directly related to hemolytic and inflammatory influencers in SCD.