New Concepts in Hemostasis
1. Hemostasis is cell-surface based. It proceeds on the negatively charged phospholipid surfaces of cells (particularly platelets) or microparticles released from platelets, leukocytes, and endothelial cells. It does not occur in the fluid phase. Anionic phospholipids or phosphatidylserine (PS) on cell surfaces provide a binding site for the assembly of coagulation factors, localizes coagulation factors, and protects them from inhibition.
2. Under physiologic conditions, the coagulation cascade is initiated by tissue factor (TF) (extrinsic pathway) and amplified by the intrinsic pathway (factor XI down). There is no role for contact pathway factors (FXII, high molecular weight kininogen and prekallikrein) in activation of the coagulation cascade. However, in pathologic arterial thrombosis (particularly that affecting the arterial system), FXII contributes to clot formation.
3. Microparticles, tiny (< 1 um) membrane-bound cytoplasmic fragments shed from leukocytes, platelets and endothelial cells, are important in hemostasis. These microparticles increase the surface area on which coagulation can proceed and can bind to the surface of activated cells, thus localizing coagulation factors derived from different cells into a focal source. Microparticles are thought to be crucial in the systemic dissemination of coagulation in pathologic states.
4. Hemostatic pathways are intrinsically linked, e.g., platelets are essential for secondary hemostasis, thrombin generated by the coagulation cascade activates platelets and inhibits fibrinolysis.
Participation of Platelets in the Coagulation Cascade
The phospholipid membrane is maintained in a state of asymmetry. Active enzymes (scramblase, floppase, translocase) ensure that the outer membrane is enriched in phosphatidylcholine and sphingomyelin, whereas the inner membrane is enriched in PS. When platelets are activated, this asymmetry is lost and PS (used to be called platelet factor 3) is exposed on the outer surface of the membrane. Also, platelets release small vesicles from their surface during activation. These microparticles are enriched in PS. PS acts as a binding site for the "tenase" (FVIIIa, FIXa, FX) and "prothrombinase" (FVa, FXa, FII) complexes of the coagulation cascade, which activate FX and prothrombin (FVII), respectively. Thus, these coagulation factors are assembled into a complex on the surface of platelet and microvesicles, where they can amplify the coagulation cascade and generate large amounts of thrombin, which converts fibrinogen to fibrin. The importance of PS in the coagulation cascade and the link between primary and secondary hemostasis cannot be over-emphasized. If there is an abnormality in PS exposure on activated platelets, the complexes cannot assemble efficiently and fibrin formation is deficient. This occurs in German Shepherd dogs and is called Scott Syndrome. These dogs suffer from severe bleeding episodes, however all routine hemostasis assays (PT, aPTT, fibrinogen, platelet count) are normal, because they do not require PS for their assessment. Only specialized tests that depend on PS exposure can detect this hemostatic abnormality (e.g., annexin V binds to PS on the surface of activated platelets and can be quantified by flow cytometry).
Activation of the Coagulation Cascade
In the past, coagulation was thought to proceed through a cascade or waterfall effect, where inactive enzymes become activated in a sequential manner, with each step being amplified. Coagulation was thought to be activated through two separate pathways, the extrinsic pathway involving TF (tissue thromboplastin or factor III) and FVII, and the intrinsic pathway, involving activation of FXII. However, this waterfall model is quite simplified and does not take into account inter-relationships between coagulation factors within the cascade (e.g., thrombin activates FIX, FVIII and FV) and between coagulation factors and cell surfaces (i.e., PS). Also, it is now known that FXII has no role in the initiation of physiologic hemostasis; the coagulation cascade is initiated by TF and the extrinsic pathway of coagulation (Figure 1). TF is a transmembrane protein that is constitutively expressed in fibroblasts in vessels, where it initiates coagulation upon injury. Small amounts of TF are found circulating in the blood; this TF is associated with microparticles, likely derived from leukocytes (particularly monocytes) and/or endothelial cells.
When a blood vessel is injured, the extracellular matrix is exposed. Circulating platelets adhere to the matrix, via vWf, and become activated, exposing PS. Also, the coagulation cascade is simultaneously initiated when TF-bearing fibroblasts in the matrix bind to plasma FVII. The TF-VIIa complex activates FX which generates a small amount of thrombin with FVa. Although TF-VIIa complex can also activate FIX directly, FX is its preferred substrate. The generated thrombin then activates FXI, FV and FVIII; FXI in turn activates FIX. The FIXa-FVIIIa-Ca2+ "prothrombinase" complex bound to PS on platelets or microvesicles then amplifies the activation of FX, which generates an explosive burst of thrombin. This large amount of thrombin then cleaves fibrinogen to fibrin and inhibits subsequent fibrinolysis, allowing fibrin to incorporate within and stabilize the platelet plug. TF-bearing microparticles become incorporated into the growing thrombus (they bind to PS-bearing platelet surfaces, thus localizing TF to activated platelet surfaces) and likely contribute to continued clot formation after initial vessel injury. Thus, the extrinsic pathway initiates coagulation, whereas the intrinsic pathway amplifies the cascade to produce sufficient thrombin to generate fibrin and prevent fibrinolysis. Fibrin is generated in two steps by thrombin. Initially, fibrinogen is cleaved to form soluble fibrin. This soluble fibrin is then crosslinked by factor XIIIa to form insoluble or crosslinked fibrin. Although primary and secondary hemostasis are linked through platelets providing PS-bearing surfaces for the coagulation cascade, they are also coupled through thrombin. Thrombin is a powerful platelet agonist, serving to recruit and activate additional platelets which become incorporated into the growing thrombus. The explosive burst of thrombin also activates an inhibitor of fibrinolysis, thrombin-activatable fibrinolytic inhibitor (TAFI).
Activation of the Fibrinolytic Pathway
Fibrinolysis is initiated when tissue plasminogen activator is activated to plasmin. This is accomplished by the release of tissue plasminogen activator (tPA) from endothelial cells. FXIIa and bradykinin (generated from FXIIa/kallikrein-mediated cleavage of high molecular weight kininogen) can also activate plasminogen and/or release tPA. Fibrinolysis is most efficient when tPA, plasminogen and c-terminal lysine residues of fibrin bind together as a complex. Plasminogen binding to fibrin markedly amplifies plasmin generation and localizes plasmin to the fibrin clot. Plasmin amplifies its own generation by exposing new lysine residues as it degrades fibrin. Plasmin is an effective, but non-specific protease. It cleaves fibrinogen, soluble fibrin and crosslinked fibrin similarly, although the degradation products released are different (Figure 2). The classic fibrin(ogen) degradation products, fragments X, Y, D and E, are yielded from plasmin action on fibrinogen or soluble fibrin. These products are detected by serum and plasma FDP assays. However, once fibrin is crosslinked, the crosslinks cannot be destroyed by plasmin so the different degradation fragments still containing the crosslinks are yielded, called crosslinked fibrin degradation products. D-dimer is a specific crosslink that is formed by FXIIIa and is exposed when plasmin cleaves crosslinked fibrin. The process of crosslinking creates a neo-epitope (new antigen), thus antibodies that react with D-dimer only detect this product in crosslinked fibrin degradation products, not in the classical fibrin(ogen) degradation products, which do not have the crosslinks. Thus, D-dimer indicates both thrombin (to activate FXIII) and plasmin (to liberate crosslinked D-dimer) generation and indicates activation of the coagulation cascade AND fibrinolysis. In contrast, FDPs can be yielded without thrombin generation (plasmin acting on fibrinogen alone) and only indicate fibrin(ogen)olysis. D-dimer testing has essentially replaced FDP testing for the diagnosis of DIC in animals. However, D-dimer is specific for fibrinolysis of any cause, either pathologic (thrombosis, DIC) or physiologic (e.g., wound healing). D-dimer test results should not be used in exclusion to diagnose DIC, but should be interpreted together with the clinical signs, patient history, and remaining hemostasis results. Titers for D-dimer should be provided, because the highest D-dimer levels (> 1000 ng/ml) will be seen in patients with thromboembolic disease.
Inhibitors are vital for keeping coagulation in check and for localizing hemostasis to the site of vessel injury. There are physiologic inhibitors for every aspect of hemostasis.
Endothelial cells release ADPase and prostacyclin, both of which inhibit platelet activation and aggregation.
A) Extrinsic pathway: The extrinsic pathway (TF-FVIIa-FXa complex) is downregulated by tissue factor pathway inhibitor (TFPI), which is mostly bound to glycosaminoglycans on endothelial cell surfaces, with small amounts found in platelets and bound to circulating lipoproteins. TFPI is released from endothelial cells by heparin and from activated platelets. There is recent evidence suggesting that protein S (a co-factor for the inactivation of thrombin) facilitates TFPI's inhibitory effect on the TF-FVIIa-FXa complex. Antithrombin also inhibits the activity of FVIIa and FXa, when bound to TF.
B) Intrinsic pathway: Thrombin binds to thrombomodulin on the surface of endothelial cells. The thrombin-thrombomodulin complex activates protein C. Activated protein C (aPC), with free protein S as a co-factor, inhibits the activation of FV and FVIII, both essential cofactors for the intrinsic pathway. With less FVa and FVIIIa, thrombin generation is decreased which reduces fibrin formation and permits fibrinolysis to proceed.
C) Antithrombin (AT) is a general inhibitor of the coagulation cascade and is activated by heparin and/or heparin-like glycosaminoglycans in the endothelial wall.
There are inhibitors of plasmin (antiplasmin) and tPA (plasminogen activator inhibitor I and II). TAFI is a fibrinolytic inhibitor that is activated by thrombin, when generated by the intrinsic pathway during amplification of coagulation. TAFI hydrolyses terminal lysine residues of fibrin, thus preventing the binding of plasminogen to fibrin, which then decreases plasmin generation.
New Concepts in DIC
1. To reiterate a well-known fact--DIC arises as a complication of underlying diseases, particularly sepsis and neoplasia.
2. The microvasculature (in which DIC occurs) is considered a distinct physiologic organ.
3. DIC is initiated by TF in most disorders.
4. Coagulation proceeds on a PS-surface--this surface is vastly increased in DIC by shedding of microvesicles (from monocytes and apoptotic cells) and lipoproteins. This facilitates the dissemination of coagulation.
5. Thrombin is pivotal in DIC--it forms fibrin clots, activates other coagulation factors amplifying its own production, activates platelets, activates (by stimulating tPA release) and inhibits (by activating TAFI) fibrinolysis, and binds to receptors on cells (protease-activated receptors or PARs), stimulating an inflammatory response.
6. Anticoagulants (AT, aPC) are crucial for limiting DIC and preventing its progression.
7. DIC is first and foremost a hypercoagulable or thrombotic disorder. Hemorrhage is a late manifestation of this syndrome.
8. Based on clinical and laboratory criteria, DIC is being separated into two distinct phases in humans: non-overt and overt DIC. Essentially, non-overt DIC reflects activation of the hemostatic system that is "challenged" but still compensated for or controlled by anticoagulants. In overt DIC, this control goes awry and hemostasis becomes uncompensated--at this stage, removal of the initiating stimulus may not halt progression. In human patients, more attention is being placed on recognition of non-overt DIC; this stage representing the best opportunity for therapeutic intervention.
There has been a recent trend towards thinking of DIC, particularly that due to sepsis, as a dysregulated response of the hemostatic system that can be divided into 4 stages: initiation, amplification and dissemination, potentiation, and endothelial dysfunction.
DIC is primarily triggered by TF, i.e., the extrinsic pathway of coagulation, in most cases through aberrant expression by monocytes and tumor cells as indicated above. In DIC, TFPI inhibition of this pathway might be ineffective.
Amplification and Dissemination
Thrombin amplifies its own expression through the intrinsic pathway as discussed above. Thrombin generation is slowed by aPC and AT. This stage of DIC, where coagulation is limited by anticoagulants, is termed non-overt DIC. The depletion of anticoagulants and the exposure of PS on cell surfaces and shed microvesicles (which circulate widely and are not rapidly cleared) results in coagulation proceeding unchecked throughout the vasculature and overt DIC.
Potentiation via Links Between Coagulation and Inflammation
A vicious cycle is initiated between coagulation and inflammation. Inflammation (especially that due to sepsis) is one of the main causes of DIC. Inflammatory cytokines induce monocytes to express TF and shed TF- and PS-rich microvesicles, which activates and amplifies coagulation, respectively. Inflammation contributes to dissemination by decreasing anticoagulants (cleaved by neutrophil proteases, decreased activation through cytokines downregulating thrombomodulin expression, and cytokine-mediated reduced hepatic synthesis). Yet inflammation is also fueled by coagulation. Thrombin binds to PARs on endothelial cells, stimulating them to release inflammatory cytokines and upregulate adhesion molecules, thus contributing to the inflammatory milieu in DIC and promoting leukocyte-mediated tissue injury. Deficiency of aPC potentiates this inflammatory state.
Endothelial cell dysfunction: When the normal adaptive mechanisms that operate to limit hemostasis to a localized site are overwhelmed, the endothelium itself becomes dysfunctional, which clinically manifests as thrombosis or hemorrhage.
Non-Overt Versus Overt DIC
The International Society of Thrombosis and Haemostasis (ISTH) has developed a scoring scheme to classify human patients into non-overt and overt DIC. This system has just begun being tested in clinical patients and initial studies suggest that it might be useful, i.e., patients in non-overt DIC have a poorer outcome (fatality rate) than those not in DIC and a greater likelihood of progressing to overt DIC and patients in overt DIC have a higher fatality rate than those not in DIC. The ISTH scoring scheme for overt DIC is primarily based on routine laboratory or "global coagulation tests" that reflect consumption or impaired synthesis, including the PT, platelet counts, fibrinogen concentration and FDP/D-dimer. This was done to make the system feasible for everyone (because other molecular tests, e.g., AT, APC are not routinely performed), and not on these necessarily being the best tests for inclusion. The distinction between non-overt and overt DIC is based on routine hemostasis assays (which may be insensitive to this phase) and more specific tests of thrombin activation (e.g., thrombin-antithrombin complexes) with depletion of inhibitors (aPC, AT). Recently, identifying trends in laboratory data over time, rather than on absolute values, has been emphasized in non-overt DIC. Veterinarians have just begun to start using the ISTH scoring schemes for DIC in dogs, but there is no consensus as to which diagnostic tests should be included in these schemes.
1. Hoffman M, Monroe DM (2001) A cell-based model of hemostasis. Thromb Hemost. 85;958-65.
2. Hopper K, Bateman SW. An updated view of hemostasis: mechanisms of hemostatic dysfunction associated with sepsis. J Vet Emerg Crit Care 2005;15:83-91.
3. Hoots WK. Non-overt disseminated intravascular coagulation: definition and pathophysiological implications. Blood Rev 2002;16 Suppl 1:S3-9.
4. Sachs UJH, Nieswandt B (2007) In vivo thrombus formation in murine models. Circ Res 100;979-991.
5. Stokol T (2003) Plasma D-dimer for the diagnosis of thromboembolic disorders in dogs. Vet Clin Small Anim 33;1419-1435.
6. Taylor FB, Jr., Toh CH, Hoots WK, et al. Towards definition, clinical and laboratory criteria, and a scoring system for disseminated intravascular coagulation. Thromb Haemost 2001;86:1327-1330.
7. Toh CH, Dennis M. Disseminated intravascular coagulation: old disease, new hope. Br Med J2003;327:974-977.