There are three pools of TFPI in the body: within microvasculature endothelial cells, stored in platelets, and circulating bound to lipoproteins.2,38 The release of TFPI from endothelial cells by both unfractionated heparin (UFH) and low molecular weight heparin (LMWH) may contribute to the efficacy of these drugs.39
TFPI is an important endogenous anti-coagulant, inactivating both FXa and the TF–VIIa complex.7,40 Initially, TFPI complexes with and inactivates FXa. The TFPI–FXa complex then binds to the TF–FVIIa complex, forming a tetramer and subsequently inactivating TF–FVIIa.2
Plasma levels of TFPI antigen have been reported to be low, normal or even elevated in septic patients.41–43 The specific role of TFPI in the development of DIC is unknown. In sepsis and the acute respiratory distress syndrome increases in plasma TF concentration without a corresponding increase in TFPI concentration have been reported.44,45
Administration of recombinant TFPI (rTFPI) in animal models of sepsis significantly reduces mortality.46–48 Despite these findings, a large phase III human clinical trial of rTFPI in patients with severe sepsis failed to reduce 28 day all-cause mortality and the risk of adverse bleeding events was higher in treated patients.49 Additional investigations are underway to investigate the therapeutic potential of rTFPI in the treatment of severe sepsis.
Anti-thrombin III, now known as anti-thrombin (AT), is a potent serine protease inhibitor produced by the liver and found in plasma and to a small extent, on the surface of endothelial cells and platelets.50 Its anti-coagulant strength is because of specific thrombin and FXa inhibition. In addition, AT can effectively inhibit the activity of factors VIIa, IXa, XIa, and XIIa. AT binds to and inactivates its target coagulation factor. This AT-factor complex is then removed by the reticuloendothelial system and as a result, the more active AT is, the faster it will be depleted.50,51
The AT molecule contains a heparin-binding domain. Glycosaminoglycans (GAGs), including heparin, heparan sulfate and dermatan sulfate, contain repeating pentasaccharide sequences, which bind to this heparin-binding domain resulting in a conformational change and activation of the AT molecule.52 This allosteric activation of AT potentiates its enzymatic activity by more than a 1000 times.50 Heparan sulfate and other GAGs are found endogenously on endothelial cells of the microvasculature and will bind to AT in the absence of exogenously administered heparin.50,51
To potentiate the inhibition of thrombin, the heparin molecule must be long enough to simultaneously attach to both thrombin and AT. This requires a heparin molecule containing at least 18 pentasaccharide units. This dual binding is not required for the inactivation of the other serine proteases.51 UFH is a heterogenous mix of molecules of varying molecular weights. A large proportion of these molecules contain more than 18 pentasaccharide sequences and as such can effectively mediate the AT inactivation of thrombin.52,53 In contrast, LMWH contains far fewer large molecules and as a consequence it has a limited ability to mediate thrombin inactivation. LMWH retains the ability to inactivate the other serine proteases, namely factor Xa.52,53 The increased FXa versus thrombin inhibitory effect explains how effective anti-coagulation with LMWH may not result in observable changes in the traditional coagulation tests such as the activated clotting time or activated partial thromboplastin time. In addition, at least in people, LMWH has a longer half-life and has far more predictable pharmacokinetics than UFH.52 As a result, a constant rate infusion of intravenous (IV) UFH in conjunction with continuous monitoring of coagulation tests can be replaced with intermittent subcutaneous administration of LMWH with minimal, if any, monitoring.53 LMWH was as effective and safe as UFH in several clinical human trials.54,55 Increased frequency of bleeding episodes is an adverse effect of LMWH administration and limits the use of this drug in patients at risk.52,53 There have been some investigations into the use of LMWH in veterinary medicine. There are reports of effective dosing of dalteparin sodiuma in dogs and horses, and there has been one investigation of dalteparin administration to cats.56–59 There have been numerous trials of several different types of LMWH in experimental animals.60–62 Each type of LMWH is unique in the composition of the heparin molecules and subsequently they have quite variable pharmacokinetics and biological properties.63 As a result, dosing information cannot be translated from one LMWH to another.
Thrombin and FXa are both pro-inflammatory in nature with wide ranging effects including increased cytokine release, leukocyte chemotaxis, and increased adhesion molecule expression.4 Hence, inhibition of thrombin and FXa by AT has significant anti-inflammatory consequences. In recent years, AT has also been found to have anti-inflammatory actions that are independent of its anti-coagulant activity.50,51 AT modulates the interaction between endothelial cells and leukocytes by reducing leukocyte activation and adhesion. AT-stimulated prostacyclin release from endothelial cells reduces platelet aggregation and decreases pro-inflammatory cytokine production.50,64,65 AT also directly inhibits interleukin-6 (IL-6) release and TF expression of endothelial cells.66 These anti-inflammatory effects rely on AT binding to endothelial cells of the microvasculature via GAGs and are virtually abolished by the administration of heparin.51,67,68 This creates a dilemma for the clinician treating a patient with heparin in order to maximize the anti-coagulant effects of AT, in doing so, the potentially important anti-inflammatory effects of AT may be lost.50,51
Both septic humans and animals have been shown to have diminished plasma concentrations of AT and in human patients, the magnitude of this reduction has been correlated with the severity of illness and survival.10,32,69,70 AT therapy in animal models of sepsis is beneficial.71,72 Despite this benefit, the phase III human clinical trial of the administration of AT concentrate to patients with severe sepsis found no effect on all-cause 28-day mortality.73 There was some evidence of beneficial effects in a subgroup of patients that were not receiving concomitant heparin therapy, although this finding should be interpreted with caution as this trial was not designed for such subgroup analysis.73 Based on this evidence AT administration was not recommended in the treatment of severe sepsis and septic shock in the latest surviving sepsis campaign guidelines.74
Protein C (PC) pathway
PC and its co-factor, protein S, are vitamin K-dependent plasma proteins synthesized by the liver. The PC anti-coagulant pathway results in the inactivation of factors Va and VIIIa, enhanced thrombin inactivation, and has direct anti-inflammatory activity.1 PC requires activation by thrombomodulin (TM), an endothelial cell membrane protein. TM binds thrombin, preventing it from continuing to activate coagulation factors, platelets and endothelial cells.75 Thrombin bound to TM is more rapidly inactivated by inhibitors such as AT, than free thrombin. The TM–thrombin complex is approximately 1000 times more effective at activating PC than TM alone.76 The endothelial cell PC receptor (EPCR) binds PC and concentrates it around the TM–thrombin complex on the surface of the endothelial cell, enhancing PC activation approximately 20-fold.77 Once activated, PC dissociates from the EPCR, binds to protein S, and this complex inactivates FVa and FVIIIa.78,79
The complex interactions involved in PC activation are an essential aspect of normal hemostasis. The microvasculature has a very large endothelial cell surface area; as a result, almost all thrombin which escapes from local sites of coagulation into the systemic circulation will be rapidly bound by TM, enhancing thrombin inactivation and potentiating PC activation.17 At the local level, factors Va and VIIIa are vital in the propagation phase of TF-mediated coagulation; their inactivation by activated PC (aPC) is a potent AT mechanism.6 PC activation is largely dependent on the TM–thrombin interaction. This is an intriguing design because thrombin, one of the most pro-coagulant substances in the body, also participates in the formation of a primary anti-coagulant substance. This physiologic design is sometimes referred to as the ‘thrombin paradox’.80 The significance of the PC pathway is demonstrated by the severe and often fatal thrombosis (purpura fulminans) manifested in individuals with an inherited PC deficiency.81
The aPC pathway has significant anti-inflammatory actions, indirectly by the inactivation of thrombin by TM and the inhibition of further thrombin generation, as well as directly.78 The EPCR has cell signaling functions and experimentally blocking the interaction between PC and EPCR has pro-inflammatory and pro-coagulant consequences.78 Soluble EPCR can be released from the endothelium in disease states such as sepsis and interferes with leukocyte–endothelium adherence.82 aPC has anti-apoptotic effects on endothelial cells via binding to EPCR and the subsequent initiation of cell signaling.83 aPC inhibits TNF release, nuclear factor-κB activation, leukocyte adhesion, and TF expression.84,85 The signaling pathways responsible for these actions of aPC are yet to be elucidated.78,79 In addition to these anti-inflammatory actions, aPC improves fibrinolysis in disease states by inhibition of plasminogen activator inhibitor-1 (PAI-1), an important mediator of impaired fibrinolysis.86,87
Numerous human studies have documented that PC is consistently depleted in sepsis.13,70,88,89 In a report of 20 septic dogs with naturally occurring sepsis, PC activity was significantly lower than controls. No relationship between PC levels and outcome could be determined.10 A study evaluating 34 septic foals also reported a significant decrease in PC activity in comparison with controls.69 PC deficiency is inversely correlated with mortality rate in human patients with severe sepsis.19,70,88 Depletion of PC during sepsis is thought to be the combined result of degradation by neutrophil elastases, consumption by increased pro-coagulant processes, inadequate hepatic biosynthesis, and possible acquired vitamin K deficiencies.76,79 The PROWESS trial, a phase III human clinical trial of recombinant aPC (Drotrecogin alfab) administration to patients with severe sepsis, resulted in a 6.1% absolute reduction in 28 days, all-cause mortality and a 19.4% reduction in the relative risk of death.18 There was an increased risk of serious bleeding found in the treated group, although serious bleeding occurred primarily in patients predisposed to hemorrhage (gastrointestinal ulceration, prolonged bleeding times, low platelet counts). Administration of aPC reduced the activation of thrombin and decreased inflammation, as determined by significant decreases in serum IL-6 concentrations in the treated patients.18 aPC is the first pharmacological agent proven to reduce mortality in severe sepsis. Administration of aPC was associated with a reduction in relative risk of death in patients regardless of the presence of an aPC deficiency prior to therapy. This result, in addition to the reduction in IL-6 levels, led to the conclusion that the effectiveness of aPC administration in septic patients is a consequence of its anti-inflammatory properties as much, if not more than its anti-coagulant properties.13,19,90 It serves as yet another example of the intimate relationship between the inflammatory and coagulation systems. The human recombinant aPC product has marked species specifity and requires a 15–20-fold higher dose in dogs to achieve a similar level of anti-coagulation in comparison with human subjects.91 In addition, drotrecogin alfa is rapidly eliminated in dogs and is antigenic in non-human species including primates and cannot be re-administered in animals for fear of anaphylaxis.c