Evaluation of Hemostatic Abnormalities in Canine Spirocercosis and Its Association with Systemic Inflammation
Department of Companion Animal Clinical Studies, Faculty of Veterinary Science, University of Pretoria, Onderstepoort, SA
Corresponding author: P. Pazzi, Department of Companion Animal Clinical Studies, Onderstepoort Academic Hospital, University of Pretoria, Private Bag X04, Onderstepoort, SA 0110; e-mail: firstname.lastname@example.org.
The study was performed in the Department of Companion Animal Clinical Studies, at the Faculty of Veterinary Science in the Onderstepoort Veterinary Academic Hospital, University of Pretoria, South Africa.
Abstract was presented at the congress of the European College of Veterinary Internal Medicine, September 2012.
Canine spirocercosis is caused by the nematode Spirocerca lupi and is characterized by esophageal fibro-inflammatory nodules that may undergo neoplastic transformation. No sensitive and specific laboratory assays other than histopathology have been reported to differentiate non-neoplastic from neoplastic disease.
Dogs with spirocercosis will have evidence of hypercoagulability based on thromboelastography (TEG)-derived maximal amplitude (MA); increased MA will be correlated with increased acute phase protein (APP) concentrations (C-reactive protein [CRP] and fibrinogen); increased MA and APPs will be exacerbated with neoplastic spirocercosis.
Thirty-nine client-owned dogs with naturally occurring spirocercosis and 15 sex-matched healthy controls.
A prospective comparative study evaluating TEG, activated partial thromboplastin time, prothrombin time, antithrombin (AT) activity, platelet count and D-dimer concentration, and APPs of dogs with non-neoplastic (n = 24) and neoplastic (n = 15) spirocercosis compared to control dogs.
Median MA was significantly increased in the non-neoplastic group (P <.01) and neoplastic group (P <.01) compared to the controls. Both APPs were significantly increased in the neoplastic group compared to the non-neoplastic and control groups. MA was strongly correlated with fibrinogen (r =0.85, P <.001) and CRP (r =0.73, P <.001). An MA >76 mm provided 96% specificity and 73% sensitivity for differentiation of disease state.
Conclusions and Clinical Importance
Canine spirocercosis is associated with increased TEG variables, MA and α, and decreased AT activity, which may indicate a hypercoagulable state seemingly more severe with neoplastic transformation. MA was correlated with APP in dogs with spirocercosis and can be used as an adjunctive test to support the suspicion of neoplastic transformation.
Spirocerca lupi (S. lupi) is a canine nematode infesting the esophageal wall resulting in the formation of fibro-inflammatory nodules. The nodules may undergo neoplastic transformation into osteosarcoma (60–85% of neoplastic cases), fibrosarcoma, or anaplastic sarcoma in up to 26% of dogs.[1-3] Early diagnosis of neoplastic transformation, preferentially by noninvasive means, is crucial to facilitate early and improved intervention which includes surgical excision and chemotherapy. Excluding histopathology of the nodule, no sensitive and specific laboratory assays have been identified to confirm a diagnosis of neoplastic transformation.
Neoplastic transformation has been reported to be closely linked with inflammation.[4-6] Evidence of systemic inflammation in spirocercosis includes pyrexia, leukocytosis as well as increased C-reactive protein (CRP) and IL-8 concentrations in the blood.[7-9] Systemic inflammation has been reported to result in activation of hemostasis, and subsequent alterations in the inflammatory and hemostatic variables in spirocercosis are expected. Definitions of hypercoagulability based on TEG variables have not yet been established for animals and numerous variables are used in different studies to define hypercoagulability, including maximal amplitude (MA) or global clot strength (G) alone or in combination with shortened reaction time (R), increased angle (α), or both.[11, 12] Increased MA has been reported to be indicative of hypercoagulability in studies on animals,[11, 13] and a single variable may be easier to interpret and may have greater clinical application. Evidence to support hemostatic alterations in spirocercosis includes three case reports of aorto-iliac thromboemboli that were attributed to the migration of S. lupi larvae and associated vascular injury.[14-16] Studies have reported decreased AT activity in dogs with spirocercosis, suggesting the presence of systemic hemostatic changes in this disease.[17, 18]
We hypothesized that (1) dogs with spirocercosis will have evidence of hypercoagulability based on TEG-derived variable, MA; (2) increased MA will be correlated with increased APP concentrations (CRP and fibrinogen); and (3) increased MA and APP concentrations will be exacerbated in dogs with neoplastic spirocercosis and may be used to differentiate dogs with neoplastic transformation.
Materials and Methods
A prospective clinical study was performed on 41 client-owned dogs with naturally occurring spirocercosis that were presented to the Onderstepoort Veterinary Academic Hospital. This study was approved by the Faculty Research Committee and the Animal Ethics committee of the University of Pretoria, South Africa.
All dogs included in the study were older than 6 months, weighed more than 6 kg and signed consent was obtained from owners for enrollment in the study. Upon admission, a complete history, clinical examination, urinalysis, fecal flotation, complete blood count (CBC), serum biochemistry profile, and thoracic radiographs were performed. Diagnosis of spirocercosis was based on positive fecal floatation, thoracic radiographs (at least 2 of the following radiolographic signs: a caudodorsal mediastinal mass, spondylitis of the caudal thoracic vertebrae or undulation of the lateral border of the descending aorta), or esophageal endoscopy.
Infected dogs were divided into 2 groups, non-neoplastic or neoplastic. The diagnosis of non-neoplastic spirocercosis was based on the endoscopic visualization of esophageal nodules with smooth surfaces and regression of the nodules at follow-up endoscopy, 6 weeks after treatment. Neoplastic transformation was based on histopathology with samples obtained by endoscopically guided biopsy, surgical excision or at necropsy. Fifteen healthy sex-matched dogs were included as controls.
Dogs were excluded from the study if they had received any macrocyclic lactone in the previous 6 months, showed signs of trauma or wounds or had been previously diagnosed with cardiac, hepatic, or renal disease. Any dogs showing concurrent systemic or inflammatory disease conditions not known to be associated with spirocercosis on clinical examination, hematology, serum biochemistry, urinalysis, and thoracic radiographs were excluded. In addition, if signs of concurrent disease were present after endoscopic regression of the nodule in the non-neoplastic group or concurrent disease was found at necropsy in the neoplastic group, the dogs were excluded from the study. Any previously diagnosed inherited coagulopathy or treatment with medication known to interfere with normal hemostasis (including corticosteroids of any kind, nonsteroidal anti-inflammatory drugs, aspirin, clopidogrel, or heparin products) either during hospitalization or 4 weeks before presentation also led to exclusion from the study.
After diagnosis, before any treatment, serum, 3.2% sodium citrate, and EDTA samples were collected from the jugular vein of each patient with a 21-gauge needle by careful venipuncture with minimum stasis into vacutainers. CBC was repeated if the diagnosis of spirocercosis was made on a day different from admission. Blood samples were collected in the order described above to avoid tissue factor contamination of the citrate sample. The sodium citrate tube was filled to ensure the correct anticoagulant-to-blood ratio (1 : 9). Citrated-blood was left to stabilize for 30 minutes after collection before TEG measurements, a fixed time point after sampling was chosen to minimize interassay variation. TEG was performed using a computerized thromboelastograph1 according to a previously described method using human recombinant tissue factor2 as the activator. Thromboelastograms were obtained for 90 minutes at 37°C. Recorded TEG variables included time to initiation of clot formation (R), time for the tracing to achieve a predetermined clot strength (K), speed of clot propagation (α), greatest clot strength (MA), calculated G value and degree of fibrinolysis at 30 and 60 minutes after MA was reached (Ly30 and Ly60, respectively; Fig 1).
The remaining citrated and serum samples were centrifuged at 2100 g for 8 minutes within 2 hours of collection and the serum or plasma removed, aliquoted and stored at −80°C for the determination of CRP concentration and other coagulation assays as a batch at a later date.[21, 22] Reference intervals for all assays included were available from the Clinical Pathology laboratory, but a healthy control group also was included for the hemostatic assays to minimize analytical variation.
PT and aPTT assays were performed on the ST art 4 analyzer3 using the Neoplastine CI Plus3 reagent kit for PT, and the C.K. Prest4 reagent kit for aPTT, according to the manufacturers' instructions. D-dimer assays were performed, using an immunometric flow-through principle.5 Fibrinogen assays were performed on the ST art 4 analyzer3 using the Sta-Fib 23 reagent kit. AT activity in plasma was measured on the Cobas Integra3 utilizing a thrombin-dependent chromogenic substrate assay.6 Serum C-reactive protein was determined using an automated CRP immunoturbidometric immunoassay for humans on the Cobas Integra.3,[] All assays were performed only once and calibrated according to the manufacturers' recommendations for human samples. Commercially available human control reagents for the respective assays were analyzed as internal controls together with pooled plasma from 5 healthy dogs.
Endoscopy and Histopathology
In addition to the previously mentioned standard diagnostics, esophageal and gastric endoscopy was performed under general anesthesia in all dogs except those where the owner declined endoscopy because of the radiographic findings of pulmonary metastasis before endoscopy. Endoscopy was performed before blood sample collection. The procedure was performed using a video endoscopy unit.7 Endoscopically guided biopsies were performed on any luminal esophageal nodule that was not smooth and showed any signs of ulceration or cauliflower-like appearance commonly seen in neoplastic transformation of esophageal nodules. A standard sampling procedure and anesthetic protocol was maintained for all spirocercosis dogs in an attempt to prevent artifactually affecting hemostasic variables. A follow-up endoscopy was performed on all dogs not euthanized, approximately 6 weeks after initial treatment. A standard necropsy was performed on all euthanized dogs as well as histopathologic examination of all esophageal nodules because of S. lupi.
Treatment was initiated after all samples and diagnostic tests were completed and it was based on a published treatment protocol of 400 μg/kg doramectin8 given by SC injection 2 weeks apart for 3 treatments.
Statistical analysis was performed using SPSS 209 and Graphpad Instat 3.10 The data were checked for normality using the Kolmogorov–Smirnov test. The Kruskall–Wallis test was used to determine significance across the 3 groups for each parameter. If significance was present, Dunn's multiple comparison was used as posthoc analysis. Correlation between inflammatory and hemostatic variables was determined using the Spearmann's rank correlation test. Data are presented as median and range of values. P < .05 was considered significant. P values are reported as <.05, <.01 or <.001 as per the limits of the software available.
Two of the 41 dogs admitted to the study were excluded: the first because of concurrent pyothorax, and the second because of severe thrombocytopenia secondary to septic pneumonia. The remaining 39 dogs were deemed free of any concurrent inflammatory conditions based on results of urinalysis, CBC, serum biochemistry, thoracic radiographs, and complete regression of clinical signs at the 6-week follow-up endoscopy or lack of concurrent disease based on postmortem examination. Twenty-four of the 39 dogs were included in the non-neoplastic group and 15 in the neoplastic group. No dogs received fluid therapy before blood collection. Three dogs in the neoplastic group did not have endoscopy performed. Seventeen breeds were represented with a significant difference in age between the control group and the neoplastic group (Table 1). The ratios of male : female for each group were as follows: control (6 : 9), non-neoplastic (13 : 11), and neoplastic (5 : 10) with no significant difference found among groups. Neoplastic transformation of the S. lupi nodule was confirmed with histopathology resulting in the diagnosis of 14 osteosarcomas and 1 anaplastic sarcoma. Six dogs with neoplastic spirocercosis had metastatic lesions at necropsy and 20% (3/15) of the neoplastic group had hypertrophic osteopathy.
Table 1. Age, hemostatic, and inflammatory variables for dogs with non-neoplastic and neoplastic spirocercosis and the healthy controls
Control Median (range)
Non-neoplastic Median (range)
Neoplastic Median (range)
RI, Reference interval, CM, Control median, N/A, not applicable.
Krusskal–Wallis showed significance between the 3 groups compared, P < .05.
Significance between control and non-neoplastic groups.
Significance between control and neoplastic groups.
Significance between non-neoplastic and neoplastic groups.
Compared to the control group, α and MA were significantly increased and K significantly decreased in both the non-neoplastic and neoplastic groups, consistent with a hypercoagulable state (Table 1, Fig 2). When the neoplastic group was compared to the non-neoplastic group, α and MA were significantly increased, consistent with a greater degree of hypercoagulability in the neoplastic group. Based on our laboratory reference intervals for MA, 33% (8/24) of the non-neoplastic group and 80% (12/15) of the neoplastic group had increased MA. In contrast, when the infected group medians were compared to the control median, 92% (22/24) of the non-neoplastic and 100% (15/15) neoplastic groups had increased MA. None of the dogs were hypocoagulable and none of the dogs in the control group had an MA outside of the laboratory reference interval. There was no significant difference between groups for R, LY30 and LY60. In an attempt to differentiate neoplastic from non-neoplastic dogs, a cutoff value of >76 mm was used for the MA in receiver operating curve (ROC) analysis resulting in a sensitivity of 73% and specificity of 96% for the differentiation of disease state (Fig 3).
The median CRP and fibrinogen concentrations were significantly increased in the neoplastic group compared to the non-neoplastic group (P < .01) as well as between the neoplastic and control group (P < .01; Table 1). When ROC analysis was applied to fibrinogen concentration to differentiate neoplastic from non-neoplastic dogs at a cutoff value of >4.2 g/L, a sensitivity of 80% and specificity of 75% were found (Fig 3).
Total leukocyte count (P < .01), mature neutrophil count (P < .01) as well as platelet count (P < .01) were significantly increased in the neoplastic group compared to the non-neoplastic group (Table 1), and hematocrit (Ht) was significantly decreased (P < .05). There was a significant increase in total leukocyte count (P < .01) and mature neutrophil count (P < .05) between the neoplastic and control groups, and a significant decrease between the 2 groups for Ht (P < .001).
For the infected groups, MA had significant positive correlations with CRP concentration (r = 0.73, P < .001), fibrinogen concentration (r = 0.85, P < .001; Fig 4), total leukocyte count (r = 0.42, P < .001), mature neutrophil count (r = 0.40, P < .01), and platelet count (r = 0.39, P = .014). Moreover, MA of the infected group had significant negative correlations with Ht (r = −0.60, P < .001) and AT activity (r = −0.34, P = .037).
Of the traditional hemostatic variables, only PT and AT activity showed significant differences between groups (Table 1). Compared to the control group, the median AT activity was significantly decreased in both the non-neoplastic (P < .001) and neoplastic (P < .001) groups. For PT, a significant difference was found only between the neoplastic group and the control group (P < .05). No significance was found between groups for D-dimer concentration.
Of the euthanized dogs (14 neoplastic), only 2 dogs had any evidence of thromboemboli at necropsy; both dogs had chronic infarcts in the kidney (1 unilaterally, the other bilaterally).
This study demonstrated that dogs with spirocercosis had increased α and MA and decreased K on the thromboelastogram, increased fibrinogen concentration, and decreased AT activity compared to healthy dogs. These hemostatic changes are supportive of a hypercoagulable state, and were more pronounced in dogs where the esophageal nodule had undergone neoplastic transformation. Indicators of inflammation, the APPs, total leukocyte and mature neutrophils were correlated with the thromboelastogram-derived MA. These findings are in agreement with previous studies of other inflammatory and neoplastic conditions in which hypercoagulability was reported based on the presence of thromboemboli or changes in TEG variables.[11, 12, 25]
The previously reported tendency of neoplastic spirocercosis to have an increased female ratio was also shown in this study, but sex has been reported not to influence coagulation based on kaolin-activated TEG, and the same can only be suspected with tissue-activated TEG. The effect of age on TEG variables has not been described and may be considered a confounder in this study as the neoplastic group had a median age that was significantly higher than the control group, and it is unclear whether the advanced age may have contributed to the increased MA of the neoplastic group. Although no dogs received fluid therapy before blood collection, the majority of dogs in the spirocercosis groups did undergo anesthesia or endoscopy. The effect of these procedures on coagulation is unclear and is likely to be minimal. MA has been reported to be indicative of hypercoagulability in previous studies in dogs,[11, 13] although the predictive value of MA for thrombotic events is questionable. The number of infected dogs with MA significantly higher than the control median was more than when the MA of infected dogs were compared to the reference interval. This finding is supportive of increased MA in spirocercosis in a controlled setting, but is of questionable clinical utility with established reference ranges, especially in the non-neoplastic group. The higher median MA in the neoplastic group compared to the non-neoplastic group allows for MA to be considered as an additional tool to differentiate between neoplastic and non-neoplastic disease. Although overlap did exist between the infected groups, an MA value of >76 mm achieved a specificity of 96% for neoplastic transformation, resulting in a very low false positive rate for neoplastic transformation above the cutoff value of 76 mm. One must be aware of the limitations when employing such a method for differentiation of disease state. Spirocercosis should be diagnosed before performing TEG because many other diseases also result in hypercoagulability. Moreover, if spirocercosis is confirmed, the presence of other inflammatory conditions must be excluded to interpret the MA value with confidence. The MA should never be viewed in isolation, but rather used as an adjunctive test in dogs where ambiguity exists. The sensitivity of the test is poor (73%) and may result in many false negative results for neoplastic transformation. Therefore, an MA >76 mm will help support the suspicion of neoplastic transformation, but will not exclude neoplastic transformation if the value is <76 mm. Histopathology remains the gold standard for diagnosis of neoplastic transformation of S. lupi esophageal nodules.
The concentrations for CRP and fibrinogen were significantly increased in the neoplastic group, and both showed strong positive correlation with MA. Significant correlation among MA, fibrinogen concentration, and platelet count has been reported in both normal and hypercoagulable people and dogs.[27-29] Fibrinogen has been reported to increase MA in both humans and dogs because of the fact that MA is primarily dependent on the cross binding of platelets and fibrinogen.[30, 31] Fibrinogen may be considered one of the “driving forces” behind inflammation because it directly influences the production of proinflammatory cytokines and chemokines,[10, 32, 33] and this is further supported in this study by the positive correlations among MA, total leukocyte count and mature neutrophil count—cells responsible for the production of cytokines and chemokines. The increased fibrinogen concentration in dogs with spirocercosis is believed to promote the inflammatory response, therefore fibrinogen is suspected of playing an important role in the increase seen in MA in these dogs. The significant difference in CRP concentration between neoplastic and non-neoplastic groups agrees with previous reports on spirocercosis. When the CRP concentration was compared between the non-neoplastic and control groups, no significant difference was found, which is surprising considering the inflammation that is observed histologically in the non-neoplastic nodules.[1, 34] The reason for this finding is unclear and is contrary to previous reports on CRP concentration in spirocercosis. The close correlation of both APPs with MA supports the interrelationship between inflammation and hemostasis and, although a consistent trend was found, the individual variation from the trend line was large, making prediction of exact values difficult. The ROC analysis of a fibrinogen cutoff concentration of >4.2 g/L to differentiate between neoplastic and non-neoplastic dogs had a very similar sensitivity compared to MA at a cutoff of 76 mm, but fibrinogen concentration had a lower specificity (75%). This would imply that the possibility of a false positive diagnosis of neoplastic transformation is greater utilizing fibrinogen concentration compared to MA. The cutoffs and subsequent sensitivity and specificities used in this study will vary among laboratories, and the establishment of laboratory-specific cutoffs to differentiate between disease states are likely to be more appropriate for the local population and laboratory.
The moderate negative correlation between MA and Ht found in this study supports the interrelationship between Ht and MA previously reported,[35-37] and it is likely that the interpretation of MA in the neoplastic group may have been affected by the significant decrease in Ht in that group compared to the non-neoplastic and control groups. Consequently, a decrease in Ht should be considered an important contributor to the change in MA. It is most likely, however not the only contributor because the correlations between MA and other variables, such as fibrinogen and CRP, were stronger than that between MA and Ht. Moreover, the role of other factors besides Ht affecting MA is supported by the significantly higher MA of the non-neoplastic group compared to the control group, despite the Ht between the 2 groups not being significantly different.
With the exception of AT activity, the traditional variables of hemostasis have been reported to be poor indicators of hypercoagulability,[11, 12] and also proved to be poor indicators of hypercoagulability in this study. AT is considered a negative APP and is associated with hypercoagulable states in humans, predisposing patients to thrombotic events.[17, 18, 38] The negative correlation of AT activity with MA in this study also supports the presence of a hypercoagulable state in dogs with spirocercosis.
Local and systemic inflammation associated with the esophageal nodules, aortic intimal damage caused by larval migration (as evidenced by 3 case reports describing aorto-iliac thromboembolism[14-16]), subsequent aortic aneurysms, and turbulent blood flow are all likely to contribute to hypercoagulability and the possible development of thrombosis in spirocercosis. No clinical suspicion existed for the presence of thromboemboli in the infected dogs of this study; however, 2 of the 14 dogs euthanized had renal infarcts as evidence of the presence of thromboemboli, but the infarcts appeared chronic and may not have been directly related to concurrent spirocercosis. Based on the low number of reported dogs with spirocercosis-related thromboemboli, as well as the low number of dogs with signs of thrombosis at necropsy in this study, clinical thrombosis associated with spirocercosis would appear to be an uncommon clinical complication despite the presence of increased MA. An alternative hypothesis is the possibility of an artifactually increased MA, secondary to anemia and hyperfibrinogenemia (particularly in the neoplastic group), but these 2 factors alone are unlikely to result in the changes seen in MA because TEG provides a holistic view of hemostasis, and hemostatic variables other than TEG also were indicative of a hypercoaguable state, further supporting the multifactorial increase in MA in spirocercosis. Regardless, spirocercosis, particularly neoplastic spirocercosis, should be a differential diagnosis of hypercoagulable states in endemic areas. The prevalence of subclinical thrombi and the effect of anticoagulant therapeutic intervention in spirocercosis are unknown and may be areas of future investigation.
In conclusion, this study identified laboratory variables indicative of hypercoagulability in canine spirocercosis with a higher degree of hypercoagulability with neoplastic transformation. A MA of >76 mm provided a specificity of 96% and sensitivity 73% for the differentiation of disease states. Fibrinogen concentration (at a cutoff of >4.2 g/L) was not as useful to differentiate with a similar sensitivity of 80%, but specificity of 75%. Factors shown to correlate with the final MA in this study included fibrinogen and CRP concentrations, Ht, total leukocyte count, mature neutrophil count, platelet count, and AT. The indicators of inflammation were most strongly positively correlated with MA, supporting the interrelationship between inflammation and hemostasis in spirocercosis. TEG can be used as an adjunctive test to support the suspicion of neoplastic transformation of the esophageal nodule as well as to determine the overall hemostatic status of the patient and assist in the therapeutic management of spirocercosis.
The study was funded by the Faculty of Veterinary Science, University of Pretoria, South Africa.
Conflict of Interest Declaration: The authors disclose no conflict of interest.
TEG® 5000 Thrombelastograph® Hemostasis System; Hemoscope, Niles, IL
Innovin; Siemens, Marburg, Germany
Cobus Integra 400 Plus Analyser; Roche, Basel, Switzerland
Diagnostica Stago, Roche, Asnières-sur-Seine, France
D-dimer Single Test; NYCOCARD Reader, ILEX, Johannesburg, Gauteng, South Africa
Cobas Integra 400 Plus Analyser; Roche Antithrombin (A), Manheim, Germany
Storz endoscopy Unit; Karl Storz Endoscopy, Tuttlingen, Germany
Dectomax®; Pfizer, Johannesburg, South Africa
SPSS Statistics 17.0 ®Software; SPSS Inc, Armonk, NY