Partial results of this study were presented at the 5th Annual Meeting of the European Society of Veterinary Clinical Pathology, September 2006, Amsterdam, The Netherlands. The study was performed in the Koret School of Veterinary Medicine, The Hebrew University of Jerusalem, Israel.
Corresponding author: G. Segev, DVM, DECVIM-CA (Internal Medicine), School of Veterinary Medicine, P.O. Box 12, Rehovot 76100, Israel; e-mail: email@example.com
Background: Vipera palaestinae is responsible for most poisonous envenomations in people and animals in Israel. Cardiac arrhythmias were reported in a retrospective study of V. palaestinae envenomations in dogs.
Hypothesis: Cardiac arrhythmias in V. palaestinae-envenomed dogs are associated with myocardial injury reflected by increased serum concentrations of cardiac troponins (cTns).
Animals: Forty-eight client-owned dogs envenomed by V. palaestinae.
Methods: Blood sampling (serum biochemistry and cTns, CBC, and coagulation tests) and electrocardiography were performed periodically up to 72 hours postenvenomation. Cardiac rhythm strips were assessed blindly for the presence and type of arrhythmias.
Results: Serum cTn-T and cTn-I concentrations were increased in 25% (n = 12) and 65% (n = 31) of the dogs at least once during hospitalization, respectively. Arrhythmias were identified in 29% (n = 14) of the dogs. Dogs with increased cTn-T had a significantly higher occurrence of arrhythmias (58 versus 19%), and higher resting heart rate upon admission and within the following 24 hours. Dogs with increased serum cTn-T concentrations were hospitalized for a significantly (P= .001) longer period compared to those with normal serum cTn-T concentrations.
Conclusions and Clinical Relevance: Dogs envenomed by V. palaestinae appear to sustain some degree of myocardial injury, as reflected by increased serum cTn concentrations and by the occurrence of arrhythmias. The latter should alert clinicians to a potentially ongoing cardiac injury. An increase in cTn-T may be of clinical relevance and indicate a cardiac injury in V. palaestinae envenomations in dogs.
Vipera palaestinae is the most common poisonous snake in Israel, and is responsible for most clinical envenomations in people and animals in the country. The snake is very common in Jordan, Lebanon, and Syria also, and is a close relative of several European and Mediterranean viper species.1–3 In a recent study, at least 30 envenomed dogs were presented yearly to the Hebrew University Veterinary Teaching Hospital, which has an annual caseload of 7,000 dogs.3
The viper's venom contains about 30 components, 16 of which have been identified, including several proteases, hemorrhagins, amino acid esterases, phospholipases (A2 and B), and neurotoxins.4 To date, no specific cardiotoxin has been identified in the viper's venom.
The clinical manifestations of V. palaestinae envenomations in dogs are mostly local, and include swelling, edema, and hematoma formation, attributed mostly to the activity of hemorrhagin in the venom. Systemic signs, however, occur quite frequently and include tachypnea, tachycardia, and cardiac arrhythmia. The mortality rate in dogs was reported to be 3.7%.1,3 Acute deaths in human patients currently are attributed mostly to anaphylactic shock.2 In dogs and other animals, the causes of death are unclear and the importance of myocardial injury is unknown, although the viper's bite has been reported to result in myocardial necrosis in 1 dog and has been associated with a late onset of myocarditis in 2 horses.5,6 In a retrospective study of 16 fatal V. palaestinae envenomations in dogs, 56% (n = 9) and 19% (n = 3) of the dogs died within the 1st 24 and 48 hours, respectively.7 The clinical importance of myocardial injury and cardiac arrhythmia as potentially contributing factors to morbidity and mortality has not been evaluated in human beings or animals.
Cardiac arrhythmias have been reported previously after V. palaestinae envenomation in both canine and human patients.6,8 In a retrospective study, 2/16 dogs fatally envenomed by V. palaestinae presented with cardiac arrhythmias that were judged severe enough to justify lidocaine treatment.7 In another recent study of V. palaestinae envenomation in dogs, cardiac arrhythmias were detected in 6 of 16 dogs that were periodically monitored electrocardiographically during hospitalization. Only a small number of dogs in that study were continuously monitored for cardiac arrhythmias.3 These arrhythmias were mostly ventricular, and included ventricular premature complexes (VPCs), idioventricular rhythms (IVR), and paroxysmal ventricular tachycardia (PVT). In that study, 65% of 327 envenomed dogs also had moderately to markedly increased serum creatine kinase (CK) activities, mostly attributed to local skeletal muscle damage at the envenomation site.3 However, the possibility that some increased CK activity originated from myocardial injury could not be excluded, because CK is present in both myocardial and skeletal muscle myocytes and no isoenzyme assays were performed.9 Cardiac arrhythmias were also reported in 42–47% of dogs envenomed by the Eastern Diamondback rattlesnake (Crotalus adamanteus).10 Whether cardiac arrhythmias in envenomed dogs truly result from direct myocardial injury or from extracardiac pathology, similar to other systemic multiorgan diseases (eg, pancreatitis, gastric dilatation-volvulus [GDV], hemangiosarcoma), is currently undetermined.11 A specific and sensitive bioassay of cardiac damage would have to be used to resolve this question.
Cardiac troponins (cTns) regulate myocardial calcium-mediated actin-myosin interactions. Cardiac troponin-I (cTn-I) and troponin-T (cTn-T) are highly conservative proteins, have similar structure in humans and dogs, and have been found to be very sensitive and specific biomarkers of myocardial injury.12,13
The aims of this prospective study of dogs envenomed by V. palaestinae were to (1) determine the prevalence and character of cardiac arrhythmias, (2) determine whether envenomations are associated with myocardial injury by measurement of serum cTn-T and cTn-I, and (3) seek an association among occurrence, severity, and type of arrhythmia and serum cTn concentrations.
Materials and Methods
Inclusion Criteria and General Data
Client-owned dogs (n = 48) were consecutively enrolled in this prospective clinical study if they had been observed to be bitten by a snake identified as V. palaestinae, manifested clinical signs, and had neither history nor signs consistent with chronic cardiac disease or acute congestive heart failure (eg, cough, weight loss, fatigue, exercise intolerance, tachypnea, and dyspnea) before envenomation. Recorded data included the date and time of envenomation, the time lag (from envenomation to admission), demographic data, progression of clinical signs during hospitalization, laboratory findings, hospitalization duration, administered medications, complications, and outcome.
Upon admission, blood samples for CBC (48 dogs) and prothrombin and activated partial thromboplastin times (PT and aPTT, respectively, 32 dogs) were collected in EDTA and trisodium-citrate tubes, respectively. These were analyzed within 30 minutes of collection with an automatic particle analyzer calibrated for canine blood.a,b Blood for serum biochemistry, upon admission (41 dogs) and at later time points, was collected in tubes containing no anticoagulants and centrifuged within 30 minutes of collection. Sera were separated, refrigerated (4 °C), and analyzed within 24 hours of collection with a wet chemistry autoanalyzer,c including alanine aminotransferase (ALT), alkaline phosphatase (ALP), albumin, amylase, aspartate-aminotransferase (AST), total calcium, cholesterol, creatinine, CK, γ-glutamyltranferase, glucose, phosphorus, total protein, total bilirubin, triglycerides, and urea. Electrolyte analysis (37 dogs), including sodium, potassium, ionized calcium, and ionized magnesium, was performed with an ion-selective electrode electrolyte analyzer.d Blood samples for cTn-T and cTn-I were collected upon admission and at 12, 24, 36, and 48 hours (in 42, 10, 28, 17, and 16 dogs, respectively) postenvenomation in tubes containing no anticoagulants, allowed to clot, and centrifuged within 30 minutes of collection. Serum samples for cTn-T and cTn-I concentrations at 72 hours postenvenomation were available from 3 and 2 dogs, respectively. Sera were stored (−80 °C), allocated, and analyzed in a single run on day 190 after the 1st collection. Analyses of cTn-T and cTn-I were performed with 2 different analyzer systems,e,f 1 of whiche has been extensively used in several previous studies of dogs.14–18 Serum cTn concentrations from the envenomed dogs were compared with reference ranges that were established for 10 clinically normal (staff-owned) control dogs.
A 6-lead ECGg was recorded upon admission and at 8, 16, 24, 36, 40, and 48 hours postenvenomation (36, 23, 29, 29, 24, 14, and 7 dogs, respectively). Additional lead II rhythm strips were recorded from selected patients that were hospitalized for longer periods at 56, 64, 72, 88, and 96 hours postenvenomation (4, 2, 3, 3, and 3 dogs, respectively). Recordings were performed in right lateral recumbency, at 25 and 50 mm/s and standard calibration (10 mm/mV), with no filtering, for a period of 60–120 seconds. All ECG data measurements and interpretations were performed in a blinded fashion (to animals, clinicopathologic results, and time points) by a board-certified veterinary cardiologist (DGO). The heart rhythm was characterized in each reading, and when judged as abnormal, its severity was further defined based on its subjectively assessed electrical instability score by means of a semiquantitative scale of 0–5, where 0 was the most mild and 5 was the most severe (Table 1). When several types of arrhythmia or conduction aberrancy were documented in a dog, even if not concomitantly, during the observation period, the final arrhythmia severity score was rendered 1 step higher than that assigned to the most severe among them. Severity scores for the entire observation period were assigned to individual patients, representing our cumulative impressions regarding electrical instability as recorded from them throughout hospitalization rather than to any given rhythm strip of any individual patient at any specific time point.
Table 1. Arrhythmia-severity scoring based on the subjectively assessed electrical stability of each rhythm (where 0 represents the most benign and 5 represents the most malignant rhythms seen in this study).
Type of Rhythm
A final severity score of each patient is based on cumulative scores of all rhythm strips recorded from that same patient throughout its hospitalization (see text and Fig 2).
Premature complexes at any HR and of any number of consecutive complexes <4, as long as an R-on-T phenomenon is not recorded during consecutive VPCs; HR ≤160/min during IVR or parasystole
HR >160/min with a sudden onset and offset, lasting >3 consecutive cycles but ≤30 seconds
Any occurrence of an R-on-T phenomenon during consecutive VPCs at any amount or rate, or a sudden onset and/or offset of an HR >160/min lasting ≤30 seconds
Sustained SVT or VT
HR >160/min lasting >30 seconds, as long as the SVT onset/offset is sudden rather than gradual
Antivenomh (26 dogs) was diluted (10/500 mL .9% saline) and administered after a hypersensitivity skin test (0.5 mL antivenom injected intradermally in a shaved area on the trunk) as a slow IV infusion to decrease the risk of anaphylactic reaction. Diphenhydraminei (2 mg/kg IM q8h, in all dogs) was administered only after the completion of the skin test and before antivenom administration. Other medications included ampicillinj (25 mg/kg IV q8h in all dogs) and various glucocorticoid preparations (9 dogs).
Descriptive statistics and statistical analyses were performed by means of 2 statistical software programs.k,l Spearman's rank correlations were performed between cTns and the severity score of arrhythmia. The Mann-Whitney U-test was used to compare median serum cTn concentrations in dogs with or without arrhythmia. A binominal proportion test was used to compare the proportion of affected dogs by sex. Fisher's exact test was used to compare the occurrence of arrhythmia in dogs with increased versus normal serum cTn concentrations and to compare the occurrence of hematologic, biochemical, and coagulation abnormalities between dogs with increased versus normal serum cTn concentrations. The resting heart rate (HR) of dogs with increased versus normal cTns was compared by Student's t-test. For all tests applied, differences were considered statistically significant when P≤ .05.
Forty-eight dogs admitted to the Hebrew University Veterinary Teaching Hospital between March and October 2004 met the inclusion criteria. There was no substantial difference between the number of females and males (27 and 21 dogs, respectively). The median age was 60 months (range, 6–144). Twenty-eight were mixed breed dogs (58%), whereas the rest were of various pure breeds. The median body weight was 25 kg (range, 6–64). The median time lag from envenomation to admission at the Hebrew University Veterinary Teaching Hospital was 120 minutes (range, 30 minutes to 1.5 days).
Dogs were bitten on the head or neck (n = 35, 73%), on a limb (n = 10, 21%), or on the tongue (n = 1, 2%), with double bites in 2 (4%) dogs (face and limb, and tongue and limb). The most common clinical signs upon admission were local swelling (n = 48, 100%), panting (n = 26, 54%), viper's fang bite marks (n = 22, 46%), hyperemic mucous membranes (n = 17, 35%), lethargy (n = 11, 23%), tachycardia (>160 beats/min, n = 10, 21%), tachypnea (>40 breaths/min, n = 10, 21%), increased body temperature (>39.5 °C, n = 8, 17%), salivation, (n = 7, 15%), local petechiae (n = 5, 10%), and dyspnea (n = 4, 8%).
CBC, Coagulation, and Serum Biochemistry Results
Most mean hematologic and coagulation parameters upon admission were within the reference intervals. Hemoconcentration was the most common abnormality observed (packed cell volume [PCV] >50%, n = 42 [88%]; >55%, n = 30 [63%], reference interval 35–50) followed by leukocytosis (>17 × 103 /mm3, n = 20 [42%]; reference interval, 5–17 × 103 /mm3) and thrombocytopenia (<200 × 103 /mm3, n = 17 [35%]; reference interval, 200–500 × 103 /mm3) (Table 2). Upon admission, mean PT (n = 32) was mildly prolonged (mean ± SD, 8.9 ± 2.3 seconds; reference interval, 6.0–8.4), whereas mean aPTT (n = 32) was within the reference interval (mean ± SD, 13.5 ± 4.4 seconds; reference interval, 11.0–17.4). However, the PT and aPTT were prolonged in 14 (44%) and 4 (13%) dogs, respectively (Table 3).
Table 2. Hematology results upon admission in 48 dogs envenomed by Vipera palaestinae.
IQR, interquartile range; RI, reference interval.
Table 3. Serum biochemistry results upon admission in 41 dogs envenomed by Vipera palaestinae.
The most common serum biochemistry abnormalities upon admission were hypokalemia (n = 19, 51%), increased total protein (n = 20, 49%), hyperglobulinemia (n = 19, 46%), hyperphosphatemia (n = 15, 37%), and increased blood urea nitrogen (BUN) (n = 14, 34%) concentrations, and increased activities of muscle enzymes (AST, CK; 12 [29%] and 11 [27%] dogs, respectively) and alkaline phosphatase (n = 11, 27%) (Table 3). Serum CK activity was above the reference interval in 88–90% of the dogs at 12–48 hours postenvenomation. Serum AST activity was above the reference interval in 90, 93, 70, and 82% of the dogs at 12, 24, 36, and 48 hours postenvenomation, respectively. Serum ALT activity was above the reference interval in 17–20% of the dogs at 12–48 hours postenvenomation.
The reference ranges for cTns established in 10 control dogs were 0–0.02 ng/mL and 0–0.1 ng/mL for cTn-T and cTn-I, respectively. Serum cTn-T and cTn-I concentrations were increased in 25% (n = 12) and 65% (n = 31) of the animals on at least 1 determination during hospitalization, respectively, and in 10% (n = 4; median, 0.24; range, 0.03–1.26 ng/mL) and 32% (n = 13; median, 0.16; range, 0.11–131.87 ng/mL) of the dogs upon admission, respectively (Fig 1). During hospitalization, at 12, 24, 36, and 48 hours postenvenomation, serum cTn-T concentrations were increased in 20% (n = 2; median, 1.31; range, 0.24–2.38 ng/mL), 14% (n = 4; median, 0.17; range, 0.06–0.24 ng/mL), 24% (n = 4; median, 0.14; range, 0.02–0.99 ng/mL), and 37% (n = 6; median, 0.07; range, 0.03–1.31 ng/mL) of the dogs, respectively. Serum cTn-I concentrations at these time points were increased in 40% (n = 4; median, 0.16; range, 0.16–250.00 ng/mL), 68% (n = 19; median, 0.51; range, 0.18–78.14 ng/mL), 64% (n = 11; median, 0.63; range, 0.18–26.76 ng/mL), and 81% (n = 13; median, 0.87; range, 0.18–35.54 ng/mL) of the dogs, respectively. Serum cTn-T and cTn-I concentrations 72 hours post envenomation (3 and 2 samples, respectively) were increased. In the latter 3 dogs, increased cTn-T and cTn-I concentrations were also detected in earlier measurements. However, serum cTn-T concentration was highest at 72 hours postenvenomation in all 3 dogs, and cTn-I at that same time point was highest in 1 of 2 dogs compared with their previously measured concentrations.
Arrhythmias, conduction disturbances, or both were identified in 29% (n = 14) of the dogs during hospitalization. These most commonly observed arrhythmias included sinus tachycardia (57%), IVR (36%), atrial premature complexes (29%), VPCs (21%), and low-grade atrio-ventricular block (1st or 2nd degree only, 21%) (Fig 2). Cardiac arrhythmias during hospitalization were significantly (P= .024) more prevalent in dogs with increased versus normal cTn-T (58 versus 19%), but not among those with increased versus normal cTn-I concentrations.
During hospitalization, there was a significant (P <.001) correlation (r= .872) between the subjectively assessed final severity score of arrhythmia and maximal serum cTn-T, but not cTn-I concentrations.
The mean resting HR of all dogs at all time points during hospitalization was within the reference interval. Upon admission, however, 34% (n = 16) had tachycardia (HR >160/min). Dogs with increased versus normal cTn-T concentrations had a significantly (P <.05) higher HR upon admission, and at 8, 16, and 24 hours postenvenomation (182 versus 126, 146 versus 117, 158 versus 116, and 148 versus 110 beats/min, respectively). Mean HR also was higher in dogs with increased versus normal cTn-I concentrations at all measured time points. This difference was significant (P= .002) only at 24 hours postenvenomation (133 versus 96 beats/min).
Treatments, Hospitalization Period, and Outcome
The median hospitalization duration was 1 day (range, 0–21). Dogs with increased serum cTn-T at least once during hospitalization had a significantly (P= .001) longer (3.8 versus 1.3 days) hospitalization duration when compared with dogs with normal serum cTn-T concentrations. There was no substantial difference in hospitalization duration between dogs with increased versus normal serum cTn-I.
There were no significant differences in median serum cTn concentrations or in the occurrence of an increase in serum cTn-T or cTn-I between antivenom-treated and nontreated dogs. Similarly, there was no association among hematologic, biochemical, or coagulation abnormalities and increased serum cTn concentrations.
One dog (2%), envenomed twice (face and a front limb), died 33 hours postenvenomation. Upon admission it had marked hemoconcentration (PCV 60%), increased serum cTn-T concentration (0.037 ng/mL), and multiple cardiac arrhythmias (IVR and severe PVT). This dog was treated with 2 U of antivenom IV. Postmortem examination was declined by the owner.
Similar to previous studies, the patients in this study were mostly young large breed dogs with no sex predilection, envenomed predominately in the head and neck area, and primarily presented with local clinical signs and hemoconcentration.1,3
Increased muscle enzyme (CK and AST) activities were the most prominent serum biochemistry abnormalities during hospitalization, as had been previously reported.1,3 The increased serum concentrations of the cardio-specific biomarkers, cTn-T and cTn-I, upon admission (10 and 32% of the dogs, respectively) suggest that myocardial injury was present early in the disease course. AST and CK are present in both skeletal and heart muscle, and are used together to diagnose and follow muscle injury.9 Because serum CK activity remained increased throughout the hospitalization period in all dogs despite its short (2 hours) half-life in dogs,19 skeletal injury, myocardial injury, or both was probably an ongoing process in these patients. The presence of ongoing myocardial injury is also supported by the increase over time of both types of serum cTn concentrations (Fig 1). In dogs, ALT is present in both liver and skeletal muscle. Under normal conditions, most serum ALT activity is of hepatic origin, and thus is considered liver specific. However, when severe skeletal muscle injury occurs, ALT loses its liver specificity.9 Thus, in the absence of other clinical and clinicopathologic evidence of hepatopathy, it is highly likely that skeletal muscle lesions were the source of the increased ALT activity observed in 17–20% of the dogs throughout the hospitalization period.
Although a specific cardiotoxin has yet to be demonstrated in V. palaestinae venom, this study suggests that dogs envenomed by this viper sustain direct or indirect myocardial injury, as reflected by increased serum cTn-T and cTn-I concentrations. The association between high serum concentrations of cTns and cardiac myocyte injury with spontaneously occurring extracardiac canine diseases has been demonstrated previously in 3 of 34 (9%) dogs with babesiosis and in 4 of 85 (5%) dogs with GDV.15,16 The mortality rate and declined postmortem examination in the current cohort limited further studies of myocardial injury, because they did not allow for microscopic and ultrastructural examinations of myocardial tissue. Nevertheless, cTns have been shown to be highly sensitive and specific biomarkers of myocardial injury.13,20,21 With the introduction of 3rd generation cTns testing, characterized by an even higher sensitivity and specificity, it is now possible to ascertain the presence of myocardial lesions in patients even in the face of concurrent skeletal muscle injury, such as existed in most dogs in the present study. It has yet to be determined whether any organic myocardial injury, if present in envenomed dogs, is a result of a specific cardiotoxin in the viper's venom or reflects secondary nonspecific complications such as microscopic myocardial infarction, necrosis, or hemorrhage. Similarly, it has yet to be determined whether arrhythmias in V. palaestinae-envenomed dogs directly result from a cardiotoxic component of the venom or are secondary to extracardiac venom effects.
The presence of myocardial injury, suggested by increased serum cTn concentrations, is further (albeit indirectly) supported by the fact that dogs with increased serum cTn-T had a significantly higher occurrence of cardiac arrhythmias as well as higher resting HR during the 1st 24 hours of hospitalization than dogs with normal serum cTn-T concentrations. Moreover, there was a significant positive correlation between serum cTn-T concentrations and the subjectively assessed final severity of arrhythmia. Depending on their rate and duration, cardiac arrhythmias (including sinus tachycardia) theoretically may have a substantial detrimental clinical impact, complicating the disease course and leading to higher morbidity and mortality rates in V. palaestinae-envenomed dogs. Although no causative relationship could be tested between myocardial injury (as reflected by increased serum cTn concentrations) and the presence, prevalence, duration, and severity of arrhythmia in our study population, the envenomation-related myocardial injury that led to increased serum cTn concentrations may have played a role in, or triggered the occurrence of, cardiac arrhythmias in these same patients. If so, the presence of cardiac arrhythmia should alert clinicians to potential myocardial injury in V. palaestinae-envenomed dogs. Additional investigation is warranted to determine whether this is true in dogs bitten by V. palaestinae or other viper species.
Although some snakebite studies have described an association between envenomation and sustained myocardial injury, others have not. A recent study of 7 children envenomed by Crotalus durissus terrificus failed to show any increase in serum cTn-I, whereas concurrent muscle enzyme (CK, AST, and lactate dehydrogenase) activities were increased.22 In another study of 45 Sri Lankan viper bites (Russell's viper and Hump-nosed viper) in human patients, no increase in serum cTn-T was observed, although some patients did present with cardiac clinical signs.23 In contrast, in a prospective study of snakebites in human patients in Papua-New Guinea, electrocardiographic abnormalities were observed in 36/69 patients (52%) envenomed by the Taipan snake (Oxyuranus scutellatus canni), in 2/6 envenomed by death adders (Acanthophis spp.), and in 1 envenomed by the Brown snake (Pseudonaja textilis). Two of 24 patients that presented with electrocardiographic abnormalities had markedly increased plasma cTn-T concentrations.24 In an experimental study in mice, IM administration of the Malayan spitting cobra (Naja sputatrix) venom induced the expression of genes involved in inflammation, apoptosis, ion transport, and energy metabolism in various tissues, including the myocardium. Electrocardiography and serum cTn-T measurements supported the presence of myocardial injury in these mice.25
Increased serum cTn-I concentration was more prevalent than increased serum cTn-T concentration in the present study. Although this study was not designed to assess the sensitivity of cTn-I, this observation may result from a higher sensitivity of cTn-I, as reflected in previous studies that found canine cTn-I to be a more sensitive cardiac biomarker than cTn-T.13,15,16 This may be supported by the fact that in the present study, in many dogs with increased serum cTn-I but with normal cTn-T concentrations, the increase in the former was only mild. In addition, in contrast to serum cTn-T concentrations, increased serum cTn-I concentrations were not associated with a higher prevalence or severity of cardiac arrhythmia, nor were they related to prolongation of hospitalization. The diagnostic sensitivity of electrocardiographic findings related to ST-segment changes as indicators of myocardial infarction in human patients is considered to be much lower than that of increased serum cTns, and has been determined to be 40–60%.26,27 This is reminiscent, to an extent, of the fact that electrocardiogarphic evidence of arrhythmia in our study population appears less indicative, relative to serum cTn concentrations, of myocardial cell injury. Moreover, the mere existence of arrhythmia does not necessarily reflect organic myocardial injury, but rather may represent an extracardiac source of arrhythmogenesis. For this reason, and in the absence of histopathology in this study, our definition of cardiac injury had to rely on the presence of a measurable increase in serum cTn concentrations. It is impossible to determine whether the relatively low occurrence of arrhythmia detection in patients with increased serum cTn-I concentration resulted from a low sensitivity of electrocardiography (a less likely option), from the fact that continuous (ambulatory) ECG recordings were not included in the study, from a low specificity of cTn-I, or from a combination of any of the above.
Although less sensitive than cTn-I, an increase in serum cTn-T may be a more clinically relevant and specific marker in dogs envenomed by V. palaestinae, reflecting a more severe or extensive myocardial insult. This hypothesis should be tested further by serum cTn measurements in a large number of envenomed dogs with the addition of myocardial histopathologic and possibly ultrastructural studies. Such a large-scale study may help determine if there is indeed an association among increased serum cTn concentrations, myocardial lesions, and survival in V. palaestinae-envenomed dogs, as has been tested previously in other diseases of dogs.16,28
The present study demonstrated substantially longer hospitalization duration in envenomed dogs with increased serum cTn-T concentration. This finding may attest to the clinical relevance of myocardial injury in V. palaestinae-envenomed dogs. Thus, serum cTn-T concentration may serve not only as a biomarker of myocardial injury but also as a prognostic factor. Because there was no association between the hospitalization period duration and serum cTn-I concentration, it appears that serum cTn-T concentration may be superior for prediction of hospitalization duration compared with cTn-I in V. palaestinae-envenomed dogs. This observation is also in agreement with a recent study investigating the prevalence of increased serum cTn-I and cTn-T concentrations in dogs with GDV.10
This study is limited by the following: First, a sample size of 48 dogs may have rendered this study underpowered. Second, the prevalence of documented arrhythmias was most likely an underestimation of their true prevalence, as most dogs were monitored only intermittently (q8h) and for very short (1–2 minutes) intervals by electrocardiography, mostly during the 1st 48 hours of hospitalization. In addition, most dogs were discharged from the hospital within this time period. Continuous, longer (≥24 hours) electrocardiographic evaluation (eg, with a Holter monitor) potentially could have identified a much higher prevalence of arrhythmias. Third, more extended serum cTn measurements may have also indicated a higher prevalence of increased concentrations of cTns. This latter scenario might be supported by the fact that at 72 hours postenvenomation, all 3 dogs tested had increased serum cTn concentrations. Moreover, in these same dogs, serum cTn-T (in 3/3 dogs) and cTn-I (in 1/2 dogs) concentrations were highest at the 72-hour sample time compared with earlier samples. Therefore, the possibility that sampling at 72 hours postenvenomation for serum cTns would have yielded a higher prevalence of increased cTns should be further investigated, as has been demonstrated in a study of GDV in dogs.16 Finally, it should be emphasized that the severity score of arrhythmia used in this study was a semiquantitative one and was devised primarily to enable subjective assessment and comparison of envenomed patients in terms of overall electrophysiological stability throughout hospitalization. It was not designed for, and could not provide, an accurate reflection of their hemodynamic or electrophysiological status, or a prediction of their risk of sudden death. For these reasons, by no means is it recommended by the authors to automatically adopt this severity score system for use under other circumstances.
In conclusion, despite the limitations described above, this study demonstrates for the 1st time that myocardial injury is present in dogs envenomed by the viper V. palaestinae, although no specific cardiotoxin has been described to date in its venom. The presence of any type of tachycardia or other cardiac arrhythmias should alert clinicians to the possibility of myocardial injury in these patients, because there was a substantial association between these 2 clinical findings and increased serum cTn concentration.