Go to start page
V1.6.8 (T332, Rfa0b21b7d)
Disclaimer & Information
Show Mindmap
Poisonous animals
Cnidarians (Jellyfish, Corals and Anemones)
Venomous fish
Hymenopterans (Bees, Wasps and Ants)
Sea snakes
Terrestrial snakes
Miscellaneous animals

Haemostatic venom/poison effects

Definition: Systemic effects that arise due to the direct action of venom/poison on components of the haemostatic system:

  1. Vascular endothelial damage;
  2. Induction and inhibition of platelet functions;
  3. Inhibition of prothrombinase complex formation;
  4. Activation of intrinsic and extrinsic clotting factors and co-factors;
  5. Prothrombin activators;
  6. Fibrinogen-coagulating activity via "thrombin-like" enzymes;
  7. Inactivation of plasma protease inhibitors;
  8. Activation of plasma proteinase inhibitors: protein C activation;
  9. Fibrin(ogen)olysis.


Signs and symptoms:

  • Bleeding from injuries, in particular ones that are not located in the region in which venom/poison application occurred;
  • bleeding into the skin (ecchymosis, petechiae);
  • gingival bleeding, haematemesis, bleeding per rectum, including melaena;
  • epistaxis, haemoptysis;
  • haematuria;
  • arterial hypotension (haemorrhagic shock!);
  • acute abdomen (intra-abdominal bleeding!);
  • flank pain/renal bed sensitive to percussion (ischaemia, renal haemorrhage!);
  • focal neurological signs, meningismus, coma (intracranial bleeding!);
  • blue sclerae (anaemia!);
  • laboratory findings: see "Diagnosis of haemostatic defects".


Venom/poison-induced haemostatic defects are almost exclusively caused by bites from venomous snakes, and thus they have been most thoroughly investigated in this area of toxicology.

As with all other components of animal venoms and poisons, haemostatically active toxins also display a high degree of species specificity. This makes it difficult to interpret the results of animal experiments with regard to envenoming/poisoning in humans. Moreover, the applicability of in vivo and in vitro investigations to the clinical situation is further limited by the fact that in the laboratory, pure toxins are used, or quantities of venom/poison much greater than those encountered in real-life conditions.

The descriptions of haemostatic defects contained in this section take these problems into account wherever possible. The haemostatic defects described here are primarily those with proven clinical significance and whose relevance in human medicine has been documented by clinical observations.


Figure 5.3 The haemostatic system. Points at which haemostatically active components of animal venoms/poisons act upon the system.

Vascular endothelial damage

Viper and crotalid venoms in particular possess marked haemorrhagic activity; a number of these haemorrhagic components, named haemorrhagins, have already been isolated.

With a few exceptions, haemorrhagic toxins are proteolytic enzymes (metalloproteases). The consequence of the proteolytic specificity of these enzymes is that they selectively cleave proteins that are involved in the structure of capillary walls. Thus, for example, the most thoroughly investigated haemorrhagic toxins, the 4 haemorrhagic metalloproteases from Crotalus atrox, are proteolytically active against fibronectin, laminin, collagen type IV, nidogen and gelatines, but do not cleave collagens type I, III and V (Baramova et al. 1989). Moreover, the proteolytic specificity of haemorrhagic toxins also constitutes a species specificity. As with all toxins, this limits the generalisability of observations made with regard to a specific species.

The great majority of the numerous proteolytic enzymes that have been isolated from snake venoms have not yet been investigated with regard to their haemorrhagic activity. In the future there are bound to be other proteases that will prove to have haemorrhagic activity. However, it is certain that not all such proteases possess haemorrhagic activity. It is still not clear which factors confer haemorrhagic properties on proteases. An attempt to answer this question was undertaken by Mandelbaum and Assakura (1988), in which they investigated the immunological properties of haemorrhagic and non-haemorrhagic proteases from Bothrops venoms, which clearly differ from each other.

Haemorrhagins damage the vascular endothelium directly, in contrast to other venom components, which cause secondary functional impairment of the endothelium. The two most important mechanisms of indirect damage to the endothelium in envenoming are:

  1. the venom-induced release of pharmacologically active substances, such as histamine, kinins, slow-reacting substances, 5-hydroxytryptamine and anaphylatoxins that increase the permeability of the vessel walls and thus cause oedema (see "Autopharmacological venom/poison effects") and
  2. venom-induced haemostatic defects with the formation of fibrinogen split products that damage the vascular endothelium (see below).

Furthermore, haemorrhagic toxins degrade fibrinogen, whereby the Aα-chain of fibrinogen is preferentially hydrolysed, in contrast to thrombin, which attacks the Aα-chain and the Bβ- chain and thus produces fibrinopeptides A and B. Apart from their effect on vessel walls, the haemorrhagic toxins also possess an anticoagulative effect, which is probably caused by the degradation of fibrin.

The difficulty in classifying venom components according to their local or systemic effects is also obvious when it comes to the haemorrhagins. Whether the haemorrhagic effects of the venom remain localised or act systemically and, for example, cause damage to the vascular endothelium of the brain, lungs, kidneys or the gastrointestinal tract is primarily dependent on the route of application of the venom, the amount of venom applied and the stage of envenoming. If the venom is applied subcutaneously or intramuscularly, which is the case in most cases of envenoming, and if, in addition, the amount of venom injected is small, the haemorrhagic effects will remain localised to the area around the site of the bite. The larger the amount of venom injected, the more marked the local symptoms will be. Large, confluent blood-filled blisters can be expected. If the venom contains both coagulation-promoting and fibrinolytic components, extensive subcutaneous bleeds may occur. Erythrocyte and fluid loss may reach such proportions as to impact the circulatory system.

If haemorrhagins enter the systemic blood circulation, they can damage endothelial cells in vascular segments far from the site of the bite. If, for example, vessels in the gastrointestinal tract are affected, this can lead to extensive blood loss and haemorrhagic shock; in critical locations, such as the central nervous system, even small bleeds are sufficient to cause serious damage. If a venom fang punctures a superficial blood vessel or a varicose vein, this naturally results per se in systemic haemorrhagic activity of the venom. Venoms that contain not only haemorrhagins but also other components that impair both platelet function and the plasmatic coagulation system are especially likely to cause a complex and life-threatening clinical course in the case of systemic application of the venom. Haemorrhagins can reach endothelial cells and damage them via two paths:

  1. via the extracellular matrix of vessels if the venom diffuses into the area immediately surrounding the site of the bite from a subcutaneous depot, and
  2. via the bloodstream if the venom (i.e. the haemorrhagin fraction) finds its way into the circulation.

If haemorrhagins degrade pericapillary extracellular matrix proteins, in particular those of the basement membrane, the integrity of the endothelial lining of the capillaries is no longer maintained and bleeds occur. Two forms of damage to the endothelial cell layer are observed, which can be divided into two forms of microscopic bleeding:

  1. Widening of the intercellular bonds between the endothelial cells, whereby the endothelial cells themselves remain intact. Erythrocytes escape from the capillaries through the "gaps" between the endothelial cells. This is known as haemorrhage per diapedesis (Fig. 5.4a) (Ownby 1982). Such a mechanism has been described for the haemorrhagin effect of Trimeresurus flavoviridis (=Protobothrops flavoviridis).
  2. The endothelial cells themselves are damaged, degenerate and are eventually lysed. Erythrocytes enter the extravascular space through the defects in the vascular endothelium that thus arise. This form of bleeding is known as haemorrhage per rhexis (Fig. 5.4b) (Ownby 1982). It is caused by the venom of Vipera palaestinae (= Daboia palaestinae, McKay et al. 1970), Crotalus atrox (Ownby et al. 1978) and Crotalus horridus horridus (Ownby and Geren 1987).



Figure 5.4 Vascular endothelial damage due to haemorrhagins.

a Haemorrhagin activity per diapedesis.

b Haemorrhagin activity per rhexis.

For further details, see Fig. 5.3.


Haemorrhagic activity plays an important role in envenoming due to Bothrops jararaca (Kamiguti et al. 1991).
In the case of diapedesis per rhexis in particular there is a massive release of procoagulant substances and expression of procoagulant properties of the endothelial cells. The extrinsic path of plasmatic coagulation and platelets are activated, such that fibrin formation and platelet aggregation occur. When haemorrhagins act locally around the site of the bite, platelet-fibrin clots form in the affected capillaries. If there is systemic haemorrhagin activity that affects extensive vessel segments, a process similar to disseminated intravascular coagulation (DIC) may occur.

Neutralisation of haemorrhagic activity with the use of polyclonal and monoclonal antibodies has been demonstrated for crotalid venoms (Ownby et al. 1984, Perez et al. 1984). Antivenoms are also able to neutralise proteolytic and haemorrhagic activity (Gutierrez et al. 1985). Heterologous antivenoms display considerable capacity for cross-neutralisation (Mebs et al. 1988).
However, all experimental attempts to neutralise haemorrhagic activity confirm the clinical experience that damage occurs rapidly and is irreversible.

The fact that bleeding in a patient bitten by a snake whose venom has a haemorrhagic effect can be controlled by the administration of antivenom does not mean that antivenom has any sort of positive influence on a haemorrhagic defect that is already established. Rather, it may be assumed that correction of concurrent thrombocytic and plasma haemostatic defects allows these defects to be resolved. If haemorrhagic toxins continue to be released into the circulation from a venom depot, they will of course be neutralised by the antivenom.

Induction and inhibition of platelet function

In recent years in particular, numerous effects of components of snake venom on platelets have been discovered and described (Brinkhous and Smith 1988). Such effects are almost exclusively caused by toxins from vipers, crotalids and a small number of elapids.

To date, only a fairly general functional classification of these venom components has been possible. The following main groups are distinguished:


Inducers of platelet aggregation

These include venom components from Bitis arietans (mechanism not yet elucidated), Bothrops atrox (protease: thrombocytin), Crotalus horridus horridus (protease: crotalocytin), Crotalus durissus terrificus (non-enzymatic platelet-activating factor: convulxin), Calloselasma rhodostoma and several Trimeresurus species (non-enzymatic platelet-activating factors: aggregoserpentines), further Bothrops species, e.g. B. jararaca and B. neuwiedi (non-enzymatic platelet-activating factors: coagglutinins) and Daboia russelli (phospholipase A2, biphasic, i.e. inhibits aggregation in a second phase). Evaluation of the clinical significance of these venom components for platelet function is difficult in all cases in which thrombin-like venom components are concurrently present.

As thrombin itself induces platelet aggregation, the contribution of non-coagulative inducers of aggregation to thrombopaenia in the course of this type of envenoming cannot be evaluated in isolation. The same applies if there is concurrent systemic haemorrhagin activity. In this situation also, thrombopaenia will have multi-factorial causes.


Inhibitors of platelet aggregation

These include venom components from several Trimeresurus species (α-fibrinogenases, nucleotidases), Calloselasma rhodostoma (α-fibrinogenase), Echis carinatus, Calloselasma rhodostoma, Bitis arietans (fibrinogen receptor antagonists) and Daboia russelli (phospholipase A2, biphasic, i.e. induces aggregation in a second phase).


Documentation of the clinical relevance of these venom effects, which act on various platelet functions, in particular the induction and inhibition of platelet aggregation, is still incomplete. It has been documented for Bitis arietans. Patients bitten by this species of viper developed thrombopaenia (Phillips et al. 1973, Warrell et al. 1975).

Inhibition of prothrombinase complex formation

Formation of the prothrombinase complex, i.e. interaction between factor X, co-factor V, calcium and phospholipid membranes during the catalytic conversion of prothrombin to thrombin, is crucial in order to achieve physiologically relevant reaction rates.

Several snake venom components interfere with prothrombinase complex formation. These include phospholipases A2 and several inhibitors without detectable enzymatic activity. Even the phospholipases A2 that are involved in the inhibition of complex formation do so not by means of their catalytic properties but rather through strong binding to the substrate. This binding occurs between non-catalytic molecular domains and receptors such as platelet and other procoagulative phospholipid membranes between which there is a high affinity. This is analogous to other such bonds formed between phospholipases A2 and suitable receptors on muscle and nerve cells and erythrocytes, with the corresponding consequences.

Inhibition of prothrombinase complex formation has anticoagulative effects. Such effects have been described for crotalid and viper species. Their clinical significance has not yet been documented.

Activation of intrinsic and extrinsic clotting factors and co-factors

Snake venoms that cause activation of intrinsic and extrinsic clotting factors/co-factors induce increased thrombin production via the coagulation cascade. In this type of envenoming, thrombin, with all its diverse activities, is at the centre of the action. To a certain extent the inhibitory systems are able to inactivate (AT III/heparin) and bind (protein C/protein S) thrombin and to inhibit new production of thrombin via negative feedback (protein C/protein S). However, these inhibitory systems are saturable, so that depending on the amount of venom injected, freely circulating thrombin may occur. The result is disseminated intravascular coagulation, which in turn sooner or later induces reactive (secondary) fibrinolysis. As with all other forms of disseminated intravascular coagulation, as soon as the process is sufficiently advanced, a paradoxical situation arises in which thrombosis formation and a bleeding tendency or even manifest bleeding are present simultaneously. Entire vessel segments, in particular their microcirculatory regions, may thrombose. The consequence is ischaemic damage to the affected organs. One of the causes of acute renal failure, i.e. renal cortical necrosis, can be explained in this way (→ Renal venom/poison efects).

The other aspect of disseminated intravascular coagulation is the consumption of clotting factors and co-factors, fibrinogen and platelets. The causes are the disseminated intravascular coagulation itself and the reactive fibrinolysis it induces. Although the primary aim of the reactive (secondary) fibrinolysis is lysis of the microthrombi, generalisation also occurs in this context, as soon as the inhibitors of fibrinolytic activity are exhausted. Free circulating plasmin occurs, which then degrades fibrinogen and clotting factors and co-factors. If the haemostatic potential is exhausted at this point, bleeding is inevitable. Bleeding into the skin occurs, as well as gingival bleeding, bleeding from puncture wounds caused by medical procedures and bleeding from predilection sites such as stomach ulcers and old wounds. If the venom also contains haemorrhagins, these types of bleeds are even more likely.

The activation of intrinsic and extrinsic clotting factors and co-factors is a mechanism of envenoming developed by vipers and crotalids. Among the vipers, these types of venom components are particularly well known and well investigated in Daboia russelli (Furie and Furie 1976, Kisiel and Canfield 1981). Extensive clinical observations are available (Than-Than et al. 1987, 1988).

Among the crotalids, the venoms of Bothrops species in particular, e.g. Bothrops atrox (thrombocytin) and Bothrops jararaca, exert these types of effects on the haemostatic system (Hofmann and Bon 1987, Kamiguti et al. 1991, Nahas et al. 1979).

Prothrombin activators

Prothrombin activators occur in the venom of Australian elapids, vipers, crotalids and colubrids and play a decisive role in the course of envenoming with many species of these genera. They are classified according to their dependence on other clotting factors and co-factors for the development of activity (Rosing et al. 1988):

  1. Prothrombin activators that are not dependent on other clotting factors and co-factors (Echis carinatus, Echis coloratus, Dispholidus typus, Thelotornis kirtlandii, Rhabdophis tigrinus). 
  2. Prothrombin activators that are dependent on phospholipids and calcium ions (Oxyuranus scutellatus scutellatus, Pseudonaja textilis, Pseudonaja affinis, Pseudonaja nuchalis, Oxyranus microlepidotus).
  3. Prothrombin activators that are dependent on phospholipids, calcium ions and factor Va (Notechis scutatus, Notechis ater, Pseudechis porphyriacus, Hoplocephalus stephensi, Tropidechis carinatus, Cryptophis nigrescens (=Rhinoplocephalus nigrescens)).
  4. Prothrombin activators that are dependent on unknown co-factors (Trimeresurus okinavensis (=Ovophis okinavensis), Bothrops atrox asper (=Bothrops asper), Bothrops jararaca, Bothrops itapetiningae, Bothrops erythromelas, Bothrops casteluandi).

The activation of prothrombin to form thrombin can occur via two paths:




Prothrombin activation by the venom of Echis, Bothrops and Dispholidus leads to the formation of meizothrombin, while Notechis and Oxyuranus venoms produce prethrombin 2 and meizothrombin. Further conversion to thrombin either occurs autocatalytically or requires factor Xa, whereby it is interesting to note that numerous prothrombin-activating snake species also possess factor X activators (Denson 1976, Bradlow et al. 1980, Kornalik and Taborska 1978, Hemker et al. 1984, cited in Rosing et al. 1988).

The thrombin formed in this way exerts numerous well-known effects on the haemostatic system, which can lead to disseminated intravascular coagulation with reactive hyperfibrinolysis as soon as the inhibitory systems are exhausted.

Fibrinogen-coagulating activity via "thrombin-like" enzymes

The term "thrombin-like" is misleading, as the only thing such enzymes have in common with α-thrombin is their fibrinogen-directed activity. The subcommittee of the "International Society on Thrombosis and Haemostasis" suggested the name "thrombic proteans" in order to solve these nomenclatural problems (Blombäk 1973). However, to date this term has not been widely adopted in the literature.

Most crotalid and some viperid venoms possess fibrinogen-coagulating activity, i.e. they are able to from a fibrin clot from fibrinogen. The proteases responsible for this process cleave fibrinopeptides from fibrinogen. Three groups of proteases can be distinguished: proteases that cleave fibrinopeptide A, fibrinopeptide B or fibrinopeptides A and B (thrombin belongs to this group) from fibrinogen. The catalytic products, fibrin monomers, spontaneously aggregate to form fibrins. In nearly all cases fibrin cross-linking does not occur, as almost all so-called "thrombin-like enzymes" not only lack the other numerous activities of thrombin but also the ability to activate factor XIII. The venom of Bitis gabonica is an exception. Moreover, fibrin monomers with only one cleaved fibrinopeptide (e.g. des-AA-fibrin) have much lower capacity to be cross-linked than those that lack both fibrinopeptides (des-AABB-fibrin) (Brostad 1978). The inhibitability of the so-called "thrombin-like enzymes" also differs from that of thrombin as they are not at all inhibitable via the physiological inhibitory systems. Thus the defibrinogenation process is not inhibited by these systems. The fibrinolytic system is secondarily activated in response to the formation and accumulation of fibrin aggregates, in particular in the microcirculation. In the absence of fibrin cross-linking, breakdown of fibrin polymers into their degradation products (FSP) is extremely efficient. Thus there is no microthrombosis with the severe consequences that are known from disseminated intravascular coagulation in the proper sense, which requires thrombin itself as driving agent. Of course any therapeutic intervention that interferes with the reactive fibrinolysis will have harmful consequences by consolidating the microcirculatory disturbance.

The accumulation of fibrin split products can be enormous, with all the consequences of inhibition of plasma and thrombocytic coagulation processes and impairment of the endothelium. If the inhibitory systems (in particular α₂-antiplasmin) are exhausted in the course of secondary fibrinolysis, fibrinogen is also degraded (fibrinogenolysis). This intensifies the defibrinogenation that is already taking place via venom-induced conversion of fibrinogen to fibrin. The fibrinogen-coagulating activity of the "thrombin-like enzymes" of Calloselasma rhodostoma (ancrod), Crotalus adamanteus (crotalase) and Bothrops atrox (batroxobin) (Stocker 1990, and references in Stocker 1990) has been particularly well investigated.

These investigations are in agreement with clinical observations made in patients bitten by these snakes, which have been well documented in the literature (Calloselasma rhodostoma, Crotalus adamanteus, Bothrops atrox).

If the defibrinogenation caused by so-called "thrombin-like enzymes" is observed in isolation, i.e. if no other components are involved that substantially increase the risk of bleeding, this process could be described as "benign". Haemostasis is maintained, as the coagulation cascade remains intact to the extent that thrombin can be produced, and generally sufficient fibrinogen, and of course platelets, are available to stop bleeding.

Inactivation of plasma protease inhibitors

The plasma protease inhibitors protein C/protein S, antithrombin III/heparin, α₂-macroglobulin, α1-antitrypsin, C1 inhibitor and α₂-antiplasmin have regulatory roles in the haemostatic system, in complement activation and in the kallikrein system. They constitute approx. 10% of all plasma proteins. Within the limits of their saturability, they represent a very efficient regulatory system of plasma proteolytic activity. This is also the case if coagulative or fibrinolytic proteases of the haemostatic system are activated by snake venom. Some snake venom components, themselves proteases, are also blocked to a certain extent by these inhibitors.

However, observations suggest that certain crotalids, vipers and colubrids have developed proteases (metalloproteases) that are not blocked at all by plasma inhibitors; in fact, they actually inactivate these inhibitors.

It is not yet foreseeable, however, to what extent these venom components are of clinical relevance in cases of envenoming. Moreover, it would be difficult to prove direct inhibition of these inhibitors by venom components, as in such cases of envenoming, consumption of inhibitors occurs anyway due to complex formation with endogenous proteases.

From an evolutionary perspective the significance of such proteases is evident. If the inhibitory systems of the prey are rendered ineffective, the coagulative and/or fibrinolytic activity of the venom will be all the more efficient (Stocker 1990).

Activation of plasma protease inhibitors: protein C activation

Protein C, together with protein S, is activated physiologically via the activator complex thrombomodulin – thrombin. Protein Ca, aided by the co-factor protein Sa, inactivates factors Va and VIIIa. The inhibitory system thrombomodulin – thrombin – protein C – protein S thus has an anticoagulative action.

Some snake venom components are also able to activate protein C and by this means exert an anticoagulative effect. These include the venoms of Daboia russelli and Agkistrodon contortrix and other viper and crotalid species (Meier and Stocker 1991).

The significance of these venom components for clinical toxicology, i.e. for the course of envenoming in patients, is not yet able to be evaluated.


Fibrin(ogen)olytic activity in clinical toxicology can be divided into 3 areas:



Primary fibrin(ogen)olysis: fibrinogenases

Fibrinogenases are primarily found in the venom of crotalids, to a lesser extent in vipers and in some elapids (Naja nigricollis).

As most venoms that contain fibrinogenases also possess fibrinogen-coagulating activity, the contribution of the fibrinogenases to the defibrinogenation is not easily distinguishable.

The significance of fibrinogenases for clinical toxicology is small compared to other toxin-induced haemostatic defects, as their range of activity is very limited (high specificity) and most fibrinogenases are inhibited by plasma inhibitors (e.g. α₂-macroglobulin).

Fibrinogenases degrade fibrinogen. They are specific for Aα or Bβ chains. Most of the degradation products are not identical to those that are formed due to plasmin activity. Fibrinogenases with Aα specificity firstly cleave the Aα chain of fibrinogen and then the Bβ chain. They have been described in venoms of various crotalid genera, Agkistrodon, Crotalus and Trimeresurus. These fibrinogenases are inhibited by plasma inhibitors (proably α₂-macroglobulin) and thus have little or probably even no significance for clinical toxicology.

It is a different case with the Bβ-chain-specific fibrinogenases, which have been described in Crotalus atrox and 2 Trimeresurus species (T. gramineus and T. macrosquamatus (=Protobothrops mucrosquamatus)).

Two of the 4 known C. atrox fibrinogenases, both Bβ chain fibrinogenases, are not inhibited by plasma inhibitors and are thus one of the possible causes of the anticoagulative effect seen following C. atrox bites (Pandya and Budzynski 1984).

With regard to the effect of fibrinogenases on fibrin, fundamentally the same applies as for fibrinogen, of course provided that there is fibrin present as a substrate for the fibrinogenases, i.e. that simultaneous toxin-induced fibrin formation is taking place. As mentioned already, this is usually the case, as the majority of fibrinogenase-containing venoms also possess fibrinogen-coagulating activity.


Primary fibrin(ogen)olysis: release of plasminogen activators

Budzynski et al. (1984) confirmed the relevance of this mechanism for clinical toxicology. Observation of a patient who was bitten by C. atrox and additional experimental investigations make it feasible that there is a toxin-induced release of plasminogen activators. A component of C. atrox venom stimulates the secretion of plasminogen activators from endothelial cells. The plasminogen activators convert plasminogen into plasmin, which then cleaves fibrinogen into the fragments X, Y, D and E. Due to their anticoagulative properties, the fibrinogen split products also impair the coagulability of the blood. This is demonstrated by the relevant clotting tests.

It is probable that C. adamanteus venom possesses the same fibrin(ogen)olytic mechanism. It is possible under experimental conditions to stimulate the release of plasminogen activators from human endothelial cells using C. adamanteus venom (Kirschbaum et al. 1988). Kitchens and Van Mierop (1983) suggested that this mechanism is one of the causes of fibrin(ogen)olysis in patients suffering from systemic envenoming following a C. adamanteus bite.

In any case of envenoming in which there is fibrinogen-coagulating activity, it is not possible to conclusively prove a plasminogen activator-releasing effect of the venom solely on the basis of observations in patients. If increased concentrations of plasminogen activators are measured in such cases, these might be caused also or solely by non-toxin-induced stimulation of endothelial cells, e.g. due to venous stasis or high local concentrations of thrombin (Francis and Marder 1990). In contrast, it can be conclusively proven that the toxin-induced release of plasminogen activators causes defibrinogenation in patients bitten by adult C. atrox. In adult C. atrox the fibrinogen-coagulating activity (thrombin-like activity) is lost and the fibrinolytic activity of the venom dominates (Reid and Theakston 1978).


Secondary (reactive) fibrin(ogen)olysis

The physiological response to the formation of a platelet-fibrin clot, for example to occlude a vessel injury, is localised fibrinolysis. Its aim is to restore vessel patency simultaneously with repair of the vessel defect. Plasma inhibitors, in particular α₂-antiplasmin, prevent this localised fibrinolytic activity from becoming generalised. Intrinsic and/or extrinsic prothrombin activation and direct prothrombin activation by venom components lead to localised or systemic fibrin formation and platelet activation. The result is microcirculatory disturbance in one or more organs, as occurs in disseminated intravascular coagulation due to other causes (e.g. sepsis, hypotensive shock). In the same way, intravascular fibrin formation due to thrombin-like fibrinogenases also leads to microcirculatory disturbances. In all of these cases secondary (reactive) fibrinolysis occurs, the aim of which is to limit the damage caused by the pathological coagulative processes. Platelet-fibrin and fibrin clots are lysed in order to restore vessel patency. The most fundamental mechanism that leads to activation of the fibrinolytic system is stimulation of the endothelial cells to release plasminogen activators. Factors that stimulate this process include venous stasis, stress and thrombin. As long as the reactive fibrinolytic activity is limited to the lysis of fibrin deposits in the microcirculation, it is fulfilling a physiological purpose. Therapeutic interventions, e.g. inhibition of the fibrinolytic activity, will be harmful.

Localisation of the fibrinolytic activity is already achieved during the formation of fibrin clots via the incorporation of plasminogen and the high affinity of tPA to fibrin-bound plasminogen, and further via plasma inhibitors of plasmin, in particular α₂-antiplasmin, and of tPA, in particular PAI-1, which rapidly inactivate any plasmin and tPA that enters the circulation. Even physiologically "reasonable" localised fibrinolysis leads to measurable levels of circulating fibrin split products. These may increase to such an extent that they interfere with controlled fibrin formation. The fibrinogen-converting activity of thrombin and spontaneous fibrin polymerisation are inhibited. In addition, platelet function is impaired.

The inhibitory systems of fibrinolysis, as with those of the coagulative system, are saturable, so that fibrinolytic activity might also become generalised. Circulating plasmin then degrades fibrinogen and thus also contributes to the defibrinogenation. The concentration of circulating fibrin(ogen) split products increases further and the above-mentioned anticoagulative and platelet function-impairing effects are intensified. Moreover, the circulating plasmin also attacks fibrin that is involved in occluding wounds, such as the bite wound, venipunctures and old wounds. This results in diffuse and localised bleeding.

The spectrum of pathological manifestations is naturally broadened if a secondary generalised fibrinolytic disturbance occurs along with the primary generalised coagulation defect. If the latter primarily impairs tissue oxygenation via microcirculatory disturbances, the risk of bleeding is massively increased by the generalised fibrinolytic activity. The bleeding risk is even greater if the coagulation potential is already exhausted.

The importance of cytokines (TNF, IL-1) in the modulation of these processes is being increasingly recognised. However, their significance for clinical toxicology cannot yet be assessed.