Haemostatic function of dogs naturally envenomed by the African puffadder (Bitis arietans) or snouted cobra (Naja annulifera)

Abstract

Snake envenomations are often medical emergencies and occur regularly in dogs. Snake venom contains hundreds of enzymes, proteins and peptides that assist in paralysing, killing and digesting prey, or as a defence against predators. Multiple snake venom components affect haemostasis in the victim. Inadvertent activation of coagulation may also result from expression of large amounts of tissue factor (TF) from injured tissues at the envenomation site, especially with potent cytotoxic venoms. The purpose of this study was to investigate the haemostatic functions in dogs envenomed by two South African snakes (Bitis arietans and Naja annulifera) using thromboelastography and traditional plasma-based coagulation assays. This prospective study included 18 client-owned dogs, of which nine dogs were envenomed by African puffadder and nine by snouted cobra. Blood was collected at presentation and at 24 hours post-envenomation. Complete blood count, thromboelastography (TEG), prothrombin time (PT), activated partial thromboplastin time (aPTT), antithrombin (AT) activity and C-reactive protein (CRP) and fibrinogen (Fib) concentrations were measured. Ten healthy client-owned dogs served as controls. These dogs were presented for routine ovariohysterectomy, castration or blood donation. Haematologic and haemostatic assay results at presentation were compared between groups using ANCOVA (analysis of covariance), and results over time between the puffadder and cobra groups were compared using linear mixed models at 5% significance. At presentation, the mean TEG R-time was significantly prolonged in the puffadder group when compared to the cobra and control groups (P=0.01 and 0.05, respectively). Visual appraisal of the thromboelastograms at presentation revealed that 5/9 (56%) of puffadder-envenomed dogs had hypocoagulable thromboelastograms as was demonstrated by prolonged R-time and decreased Angle (á), maximal amplitude (MA) and global clot strength (G). Despite this observation of hypocoagulability, none of the other TEG parameters (á, MA or G) were significantly decreased when compared to the cobra and control groups. This finding of hypocoagulability was surprising, because puffadder venom is cytotoxic, often inducing severe tissue necrosis and potentially leading to limb loss and disability in people. It therefore seems that certain components in puffadder venom affect the thromboelastograph by either interfering with or consuming coagulation factors, resulting in a hypocoagulable tracing. It is also possible that this is a dose-dependent effect, with only dogs with a significant amount of envenoming demonstrating this phenomenon. This effect appears to be transient, as 6/8 dogs (one fatality) envenomed by puffadders reverted to a severely hypercoagulable state at 24 hours post-envenomation. One dog was still hypocoagulable and one dog that was hypocoagulable became normocoagulable but still had a prolonged R-time. In the cobra-envenomed group hypercoagulable thromboelastograms were observed in 5/9 (56%) dogs at presentation as was demonstrated by increased MA and G. At 24 hours post-envenomation all cobra-envenomed dogs demonstrated hypercoagulable thromboelastograms. This hypercoagulability at presentation and 24 hours post-envenoming was not statistically significant between groups. This hypercoagulable state was likely due to tissue factor-activated coagulation promoted by inflammation at the envenomation site. At presentation, marked thrombocytopenia was evident in the puffadder-envenomed dogs when compared to the cobras and controls (P=0.04 and 0.001, respectively). Thrombocytopenia following puffadder envenomation has been reported in dogs and baboons. Components have been identified in puffadder venom that interfere with platelet function either by inhibiting or promoting aggregation. At 24 hours post-envenomation mean platelet count (Plt) was mildly increased compared to its value at presentation in the puffadder-envenomed dogs. There were Plt abnormalities in the cobra-envenomed dogs at presentation or at 24 hours post-envenomation. Marked leucocytosis was detected in the puffadder-envenomed dogs at presentation when compared to the cobras and controls (P=0.003 and 0.001, respectively) and was more severe at 24 hours post-envenomation when compared to the cobra group (P=0.01). Leucocytosis has been reported in different types of snake envenoming including puffadder-envenomed dogs. C-reactive protein (CRP) concentration at presentation was below the lowest detection limit for most dogs (14/18) in this study. At 24 hours post-envenoming all but two dogs (one each in the puffadder and cobra groups) had severely elevated CRP. This increase in CRP was statistically significant in both puffadder and cobra-envenomed dogs when compared to its concentration at presentation (P=0.04 and 0.001, respectively). Fibrinogen (Fib) concentration was not elevated in any envenomed dogs at presentation, but increased 24 hours post-envenoming. Although this increase was not statistically significant, an increase would suggest activation of the inflammatory response, as both Fib and CRP are positive acute phase proteins. Elevated CRP, neutrophilic leucocytosis and increases in cytokine IL-6 and IL-8 has been documented in four human patients bitten by Bothrops and Crotalus snakes (two each) in Brazil. CRP levels were low immediately post-envenoming, peaked at two days post-envenoming and dropped to within normal limits four days post-envenoming. These findings demonstrated a typical acute-phase response and it is likely that a similar acute phase response occurred after puffadder and cobra envenoming in our study. Mean antithrombin (AT) activity was mildly decreased in both the puffadder- and cobra-envenomed dogs compared to the controls (P=0.002 and 0.004, respectively), suggesting that the activation of haemostasis led to some AT consumption. Mean PT and mean aPTT were prolonged in the cobra-envenomed dogs compared to the controls (P=0.03 for both), but were within their reference intervals (RI). At 24 hours post envenomation mean haematocrit (Ht) was significantly decreased in the puffadder group compared to the cobra group (P=0.01), but was within RI. The Ht was significantly lower at 24 hours post envenomation compared to presentation values in both these groups (P<0.001 and 0.02, respectively). At presentation, marked thrombocytopenia, leucocytosis and prolonged clot initiation were common features in puffadder-envenomed dogs and were likely venom-induced. Snouted cobra-envenomed dogs were normo- to hypercoagulable at presentation. Both puffadder- and cobra-envenomed dogs equally showed hypercoagulability at 24 hours post-envenomation and this was more pronounced compared to their coagulability at presentation. TEG proved to be a useful tool to detect abnormal haemostasis in all envenomed dogs in this study. TEG also provided additional insights into certain aspects of snake envenomation (such as hypercoagulability) that has not been reported on previously and cannot be assessed using traditional coagulation assays. TEG may serve as a differentiating tool in early envenomation between these two types of snake envenoming in scenarios where the identity of the snake species involved is not known.