Hemostatic Aspects of Sickle Cell Disease

The Red Blood Cell and Hemoglobin S Polymerization

Sickle hemoglobin (HbS) occurs when the normal β glutamic acid residue is replaced by valine (GAG to GTG mutation at codon β ). The polymerization that occurs when HbS (α ₂ β ₂ ^S ) is deoxygenated is the primary event in the pathophysiology of SCD and results in damage to erythrocytes, tissues, and organs. Notwithstanding this straightforward molecular basis, the pathophysiology of clinical disease is exceedingly complicated. The rate and extent of HbS polymer formation is dependent on the intraerythrocytic HbS concentration, the degree of hemoglobin deoxygenation, and the intracellular concentration of fetal hemoglobin (HbF). The HbS polymer is a twisted, rope-like structure composed of 14 strands that distorts the red blood cell into the classic sickle shape. The Hb tetramer is oriented such that in one of the two β subunits, β valine forms a hydrophobic contact with a complementary site on a β subunit of the partner strand. There is evidence that the polymerization of HbS is extremely cooperative and can be regarded as a simple crystal-solution equilibrium. The lag period required for the formation of polymer is referred to as the delay time. As the range of transit times in the microcirculation is short relative to the range of delay times of sickle red blood cells (RBCs), polymers do not form in most of the cells. If, however, sickle RBCs are subjected to prolonged transit times, then HbS polymer would form in almost all the cells as a result of equilibration at the lower oxygen tension. Hb F inhibits the polymerization of Hb S, primarily owing to the glutamine residue at codon γ87, which prevents a critical lateral contact in the double strand of the sickle fiber.

The density distribution of sickle RBCs is very broad, due mainly to the high number of reticulocytes with a relatively low intracellular hemoglobin concentration and the presence of a high number of very dense cells. These cells appear dense on microscopy because of enhanced cellular dehydration following polymerization-induced damage to the cell membrane. As the rate of HbS polymerization is strongly dependent on the intracellular hemoglobin concentration, dense sickle RBCs are more likely than less dehydrated cells to polymerize and contribute to the hemolytic and vaso-occlusive aspects of SCD.

The major clinical manifestations of SCD appear to be driven by two major pathophysiologic processes: vaso-occlusion with ischemia–reperfusion injury and hemolytic anemia. Acute vaso-occlusive episodes are thought to be caused by the entrapment of RBCs and leukocytes in the microcirculation, with resultant vascular obstruction and tissue ischemia. These vaso-occlusive events are usually triggered by inflammatory stimuli, which increase adhesive interactions between both RBCs and leukocytes and the endothelium in postcapillary venules, resulting in vascular occlusion. The obstruction of precapillary venules by rigid and deformed RBCs also contributes to microvascular occlusion. Microvascular occlusion and tissue ischemia are usually followed by the restoration of blood flow, which promotes tissue injury mediated by reperfusion with increased oxidant stress, inflammatory stress, and increased expression of endothelial cell-adhesion molecules.

Leukocytes

The important contribution of leukocytes to the pathogenesis of the sickle hemoglobinopathies is illustrated by the clinical findings that elevation of the leukocyte count is recognized as a risk for early death, acute chest syndrome, and hemorrhagic stroke in patients with SCA. Episodes of severe vaso-occlusive crisis and acute chest syndrome have occurred following administration of granulocyte-colony stimulating factor (G-CSF) to patients with SCD in their “steady state.” Leukocytes interact with sickle RBCs and vascular endothelium and are stimulated to release injurious cytokines. The adhesion of leukocytes to vascular endothelium is mediated by several adhesion molecules. In addition to the adhesion to vascular endothelium, leukocytes interact with platelets and erythrocytes to form cell aggregates, stabilized via CD36-TSP, CD31-CD31, and CD62L-CD162 bonds ( Fig. 41.1 ). In patients with SCD, these cell aggregates can more effectively occlude the microvasculature than single cells. Following surgical preparation of cremaster muscle of sickle cell mice for intravital microscopy, an inflammatory response leading to rolling and adhesion of leukocytes to venular endothelium is often observed, followed by interaction of RBCs with adherent white blood cells (WBCs). However, the administration of intravenous immune globulin (IVIg) has been shown to reduce both the number of adherent WBCs attached to the endothelium and the number of interactions between RBCs and WBCs in a dose-dependent manner. In addition, mice lacking both P- and E-selectins, in which leukocytes are prevented from recruitment to the endothelium, are protected from vaso-occlusion in this model.

Activated neutrophils contribute to endothelial damage. During episodes of infection, increased numbers of activated neutrophils secrete inflammatory cytokines, which activate the vascular endothelium. Furthermore, these activated leukocytes express increased levels of adhesion molecules and attach more readily to activated endothelium. Circulating neutrophils are heterogeneous, likely due to aging in their circulation and replacement by newly released neutrophils from the bone marrow. Aged neutrophils, marked by CD62L ^lo CXCR4 ^hi , are an active subset with enhanced Mac-1 activation and the formation of neutrophil extracellular traps (NETs) under inflammatory conditions. Neutrophil aging is driven by microbiota-derived signals through neutrophil Toll-like receptors (TLRs) and Myd88-mediated signaling. Aged neutrophils are increased in SCD mice, and the increased aged neutrophil counts correlate positively with neutrophil adhesion, Mac-1 activation, and neutrophil–RBC interactions. The use of broad-spectrum antibiotics to deplete microbiota decreases the number of aged neutrophils and reduces neutrophil adhesion, Mac-1 activation, and neutrophil–RBC interactions in SCD mice, with protection from tissue damage and prolongation of survival. Similar to findings in SCD mice, aged neutrophils are increased in patients with SCD compared with healthy controls. Furthermore, the number of aged neutrophils was noted to be significantly reduced in patients on penicillin prophylaxis, although it remains uncertain whether antibiotics can decrease vaso-occlusive episodes.

Monocytes may enhance vaso-occlusion in SCD by contributing to endothelial activation. Monocytes from patients with SCA are activated and can enhance vaso-occlusion through an endothelial inflammatory response promoted by the nuclear factor-kappa B-mediated upregulation of adhesion molecules and tissue factor (TF). The activation of endothelial cells by sickle monocytes appears to be mediated by tumor necrosis factor (TNF)-α and interleukin (IL)-1β, both markers of monocyte activation. The activated endothelial cells increase their expression of ligands for adhesion molecules on leukocytes and RBCs, thereby promoting vaso-occlusion.

Platelets

Older children and adults with SCD typically exhibit moderate degrees of thrombocytosis. These patients also exhibit increased numbers of young, metabolically active platelets (megathrombocytes), a finding that is attributed to a loss of splenic sequestration following the functional asplenia observed in SCA patients. Although there are conflicting reports regarding platelet survival in SCD, platelet aggregation does appear to be increased in adult patients in the noncrisis steady state, possibly due to an increase in the number of megathrombocytes in the peripheral circulation. This increased platelet aggregation could also reflect increases in the circulating levels of such platelet agonists as thrombin, adenosine diphosphate (ADP), or adrenaline. In children, however, platelet aggregation is normal or reduced, a finding attributed to preservation of some of their splenic function and/or to fewer circulating megathrombocytes. The reduced platelet responsiveness to aggregating agents observed in children with SCD may be a result of ongoing platelet activation and secretion in vivo, changes that could, in turn, cause a depletion of platelet granule stores.

There is evidence of increased platelet activation in the noncrisis steady state. Patients with SCD have decreased platelet thrombospondin-1 (TSP-1) and CD40 ligand content when compared with normal controls, suggesting a state of ongoing release and depletion of both TSP-1 and CD40 ligand from activated platelets. Platelet expression of CD62, CD63, and PAC-1 antigen in SCD patients are significantly increased compared with ethnically matched and nonmatched controls. The expression of both P-selectin (CD62P) and CD40 ligand is substantially higher in children with SCD than in healthy control subjects. In addition to elevated plasma levels of the α-granule constituents, thrombospondin, platelet factor 4, and β-thromboglobulin, platelet-derived plasma-soluble CD40 ligand and TNFSF14 (LIGHT) are increased in the noncrisis steady state compared with normal controls.

Circulating platelet aggregates (including both platelet-erythrocyte and platelet-monocyte aggregates) are increased in SCD patients during the noncrisis state and appear to increase further during acute pain episodes. Platelets have also been detected in the heterotypic synergy between the monocyte and reticulocyte in a P-selectin/P-selectin glycoprotein ligand-1 dependent interaction.

Platelet procoagulant activity has been reported to be significantly increased in patients during acute pain episodes compared with the noncrisis state and was significantly correlated with the number of pain episodes the following year. A higher level of soluble CD40 ligand was also reported in patients with more frequent pain episodes, although the difference was not statistically significant. Platelet activation assessed by the activated fibrinogen receptor glycoprotein IIb/IIIa is correlated with echocardiography-derived tricuspid regurgitant jet velocity and laboratory markers of hemolysis. Furthermore, administration of sildenafil, a phosphodiesterase-5 inhibitor that potentiates nitric oxide (NO)-dependent signaling, has been shown to decrease platelet activation.

Endothelium

As the potential for sickle RBCs to initiate a vaso-occlusive event is dependent on whether the rate of polymer formation is within the range of the capillary transit time, any factor that slows the transit of sickle RBC through the microcirculation could be expected to have an effect on the pathogenesis of vaso-occlusion. The degree of adherence of sickle RBCs to vascular endothelium strongly correlates with the severity of disease. Multiple studies in static and dynamic conditions demonstrate that sickle RBCs attach more readily to cultured endothelial cells than do normal RBCs.

These adhesion reactions are mediated mainly by interactions between receptors on WBCs, sickle RBCs, and endothelial cells, or subendothelial matrix proteins. Although the adhesion of leukocytes to the endothelium during inflammation can involve multiple molecules, the process is initiated by P-selectin. The expression of P-selectin on the endothelial surface mediates abnormal rolling of leukocytes and static adhesion of sickle RBCs to the vessel surface in vitro. The plasma ligand thrombospondin (TSP) provides a bridge between the RBC receptor CD36 and several constitutively expressed endothelial receptors. _. Because TSP comprises a number of heterogeneously distinct domains, vascular adhesion to TSP depends on several endothelial sites, including the vitronectin receptor (α _v β ₃ ), the transmembrane glycoprotein CD36, and endothelial cell surface heparan sulfate proteoglycans. Sickle RBCs also interact with immobilized TSP via the integrin-associated protein CD47, a molecule associated with the Rhesus membrane complex. Interactions occur between the integrin complex, α ₄ β ₁ (VLA-4), expressed on reticulocytes, and both endothelial vascular-cell adhesion molecule-1 (VCAM-1) —a molecule expressed on the surface of endothelial cells (especially following activation by inflammatory cytokines and hypoxia)—and fibronectin. High-molecular-weight multimers of von Willebrand factor (vWF) promote red cell adhesion to endothelial vitronectin receptor α _v β ₃ and the GPIb-IX-V complex. Interactions also occur between sickle RBCs and subendothelial immobilized matrix proteins, including laminin, TSP, vWF, and fibronectin, proteins also present in plasma in a soluble form. Laminin binds avidly to sickle RBCs via the erythrocyte basal cell adhesion molecule-Lutheran protein receptor (B-CAM/LU), the protein that carries Lutheran blood-group antigens. Non-receptor-mediated adhesive mechanisms include a role for RBC sulfated glycolipids and phosphatidylserine (PS).

Inflammation

SCD is often referred to as a chronic inflammatory state owing to the presence of a chronic elevation in leukocyte counts, shortened leukocyte half-life, and abnormal activation of neutrophils and monocytes. The circulating endothelial cells in patients with SCD are activated with proadhesive and procoagulant properties and exhibit evidence of oxidative stress. There is also evidence for activation of the coagulation system, with activation of circulating platelets, increased number of microparticles (MPs) derived from RBCs, platelets, monocytes, and endothelial cells. In addition to these observations, SCD patients, even in the noncrisis “steady state,” exhibit elevated levels of inflammatory mediators (such as IL-6, TNF-α, IL-1, and placental growth factor), acute phase reactants (such as C-reactive protein, secretory phospholipase A2, and G-CSF), and markers of endothelial cell injury (such as soluble VCAM-1). The inflammatory biology in patients with SCD may result from infection as well as sickle RBC adhesion to endothelium, the reperfusion-injury physiology observed in these patients, and hemolysis.

Invariant natural killer T cells (iNKT cells), a subset of T cells, play a key role in promoting pulmonary inflammation and dysfunction in SCD. More numerous and activated iNKT cells (CD69 ⁺ interferon [IFN]-γ ⁺ ) hypersensitive to hypoxia/reoxygenation are found in the spleen, liver, and lung of NY1DD transgenic SCD mice compared with controls. Furthermore, there are more numerous and activated iNKT cells in the circulation of patients with SCD than in control subjects. During painful episodes, iNKT cells become more activated and express higher levels of A _2A R in an NF-κB–dependent manner.