Evaluation of hemostasis in hyperthyroid cats

Abstract Background Hyperthyroid cats might have a predisposition to arterial thrombus formation. The mechanism for thrombogenesis currently is unknown but could be associated with systemic hypercoagulability as seen in hyperthyroid humans. Objective Our purpose was to evaluate markers of hemostasis in hyperthyroid cats compared to healthy cats, and in hyperthyroid cats before and after radioactive iodine treatment (RIT). Animals Twenty‐five cats with hyperthyroidism and 13 healthy euthyroid cats >8 years of age. Methods Prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen concentration, antithrombin (AT), D‐dimers, thrombin‐antithrombin complexes (TAT), von Willebrand Factor antigen (vWF : Ag), and activity of factors VIII and IX were measured. An echocardiogram was performed in all cats. Hemostatic markers and echocardiogram were evaluated again 6 to 9 months after successful RIT in 7 cats. Results Hyperthyroid cats had higher fibrinogen concentration (P < .0001), AT activity (P < .0001), and vWF : Ag concentration (P = .01) than healthy control cats with all results decreasing significantly post‐RIT. Hyperthyroid cats were not more likely to be in a hypercoaguable state than euthyroid cats (P = .08). Serum T4 concentration was not a predictor of a hypercoagulable state (P = .53). Conclusions and Clinical Importance Hyperthyroid cats have evidence of altered hemostasis that does not appear to be solely attributable to cardiac abnormalities, but no evidence of a hypercoagulable state. Findings suggest altered hemostasis resolves after RIT. Hyperthyroid cats could have endothelial dysfunction as indicated by increased vWF : Ag which could potentiate thrombogenesis.


| INTRODUCTION
In humans, hyperthyroidism alters the coagulation-fibrinolytic balance creating a hypercoagulable state that increases the risk of thromboembolism (TE). 1 The risk of venous thrombosis increases with higher serum concentrations of free thyroxine in people. 2 The mechanisms by which hyperthyroidism induces a hypercoagulable state in humans are thought to involve all stages of hemostasis. Primary hemostasis is enhanced by increased circulating von Willebrand Factor (vWF) concentration, which promotes platelet adhesion to the subendothelium. 3,4 Hyperthyroidism also induces increased secondary hemostatic function by increases in coagulation proteins such as factor VIII, factor IX, and fibrinogen. 3,4 In addition, fibrinolysis is impaired as a result of increased plasminogen activator inhibitor concentration. 3,4 Finally, hyperthyroidism also leads to endothelial dysfunction, which contributes to a hypercoagulable state in people. 4 Similar to hyperthyroid humans, cats with hyperthyroidism might be in a hypercoagulable state, but the mechanism of potential thrombus formation in hyperthyroid cats is unknown. The patterns of thrombosis in the 2 species are different. Venous thrombosis is described in hyperthyroid people, whereas arterial thrombosis is most common in cats. 1,5 Mechanisms of thrombus formation are multifactorial and include abnormal blood flow, endothelial damage, and hypercoagulability. 6 In cats with cardiomyopathy, thrombus formation has been attributed to blood stasis and endothelial injury. 7,8 Some evidence suggests that cats with cardiomyopathy also might be hypercoagulable. [8][9][10][11] Hyperthyroidism results in complex changes that alter circulation and cardiac structure and function. 11 The resultant thyrotoxic cardiac disease could contribute to thrombus formation in hyperthyroid cats. 12,13 However, a hyperthyroid cat with an echocardiographically normal heart has been documented to have had an episode of arterial thromboembolism (ATE). 5 A predisposition for thrombus formation can exist independent of structural abnormalities such as atrial enlargement. 10 This possibility is consistent with the existence of mechanisms other than cardiac disease that might be responsible for TE in hyperthyroid cats. 5 Hyperthyroid cats have been documented to have changes consistent with a hypercoagulable state as indicated by hyperfibrinogenemia and shortened PT, which might explain the occurrence of TE in affected cats that do not have clinically detectable thyrotoxic heart disease. 14 The purpose of our study is to evaluate hemostatic markers in hyperthyroid cats as compared to healthy controls and after treatment with RIT. Our hypotheses were that hyperthyroid cats would have altered hemostatic markers and that altered hemostasis would resolve after successful RIT. We also hypothesized evidence of systemic hypercoagulability would be found.
(PSI-70BT, Toshiba Medical Systems, Tokyo, Japan). All measurements were performed off-line by a single operator (JA), blinded to the cat's thyroid status. End-diastolic left ventricular dimensions were measured on M-mode images obtained with 2-dimensional (2D) guidance from a right parasternal short-axis view of the left ventricle at the level of the papillary muscles. End-diastole was defined as the onset of the QRS complex on the concomitantly acquired ECG tracing, and measurements were performed using "leading edge-to-leading edge" technique. Additionally, right parasternal long-axis images in which the outflow tract was visible were analyzed. End-diastolic frames were identified as the first frame of closure of the mitral valve. In these frames, the thickest segment of the interventricular septum was measured using "leading edge-to-trailing edge" technique, perpendicular to the endocardial borders; measurements of the left ventricular (LV) lumen were made perpendicular to the long-axis of the ventricle at the thickest region and excluded the endocardium; and, the thickest portion of the posterior wall was measured using "leading edge-toleading edge" technique, perpendicular to the endocardial border and the pericardium. Papillary muscles were excluded from all measurements. From short-axis images, left atrial and aortic dimensions were Hyperthyroid cats received RIT by SC route, with dosing based on the patient's serum T4 concentration and thyroid scintigraphy results and, if documented to have resolution of hyperthyroidism as defined by a serum T4 concentration within the reference interval 6 to 9 months after treatment, had an echocardiogram performed identical to that before treatment. All echocardiograms were performed by the same operator and measurements were made by an observer blinded to the status of the cat as described above.

| Statistical analysis
Statistical analyses were performed by commercially available computer software (SAS Version 9.4, Cary, North Carolina). A power analysis based on previous findings showed that 13 cats per group (hyperthyroid and healthy controls) would be needed to detect a difference of 0.40 in the proportions of cats in a hypercoagulable state (defined below) with a power of 82.8%. 8 The proportion of cats with a hypercoagulable state was compared between the 2 groups (hyperthyroid and control) by Fisher's exact test.
Predications for the prevalence of hypercoagulability were made based on a previous study evaluating hypercoagulability in cats with cardiac disease. 8 For continuous variables, normal probability plots were used to determine whether the distribution of the data was approximately normal. Normal probability plots showed that all coagulation variables including PT, aPTT, fibrinogen, AT, D-dimers, vWF : Ag, TAT, factor VIII, and factor IX were not normally distributed. The corresponding data were summarized as medians (range). The Wilcoxon rank sum test then was used to compare coagulation variables (1 at a time) between groups of cats defined as follows: (a) initial hyperthyroid cats versus controls; (b) initial hyperthyroid cats with normal echocardiograms versus cats with abnormal echocardiograms; and (c) cats that required sedation for blood collection versus cats that did not require sedation for blood collection.   Nine cats in the hyperthyroid group underwent a second evaluation at a median of 6 months post-RIT (range, 6-9 months). Ten of the initial hyperthyroid cats were lost to follow-up, 3 were euthanized before the reevaluation time frame, 2 were hyperthyroid based on serum T4 concentration at 6 months post-RIT, and 1 did not receive RIT.

| Serum T4 concentrations
The median serum T4 concentration in the hyperthyroid cats was  other measured hemostatic markers ( Table 2). We did not identify a linear association between serum T4 concentration and LA : Ao ratio.

| Echocardiography
Seven cats in the hyperthyroid group had echocardiography performed after RIT and documentation of normal serum T4 concentration. Six of the 7 cats had abnormal echocardiograms on initial presentation. Only 1 of these 6 cats had resolution of the echocardiographic changes after treatment. Comparison of LA : Ao obtained before and after treatment did not identify a difference (P = .05).
Cats with hyperthyroidism had higher plasma fibrinogen concentration than euthyroid controls (P < .0001; Figure 2). If hyperthyroidism resolved after RIT, fibrinogen concentration decreased (P = .02; The hyperthyroid group also had higher vWF : Ag concentrations compared to euthyroid controls (P = .01; Figure 2) that decreased after treatment (Figure 3). No difference was found in D-dimer concentrations between hyperthyroid cats and euthyroid controls, but after RIT, the D-dimer concentration increased compared to initial results in hyperthyroid cats (P = .02; Figure 3).
For the remaining markers of hemostasis in hyperthyroid cats (PT, aPTT, TAT, factor VIII, factor IX), no differences were found when compared to euthyroid controls or post-RIT results (Tables 3 and 4). Sedation did not have an apparent effect on any hemostatic variable.

| Hypercoagulability
Two hyperthyroid cats were excluded from analysis of hypercoagulability because of lack of factor VIII and factor IX measurements, leaving 23 hyperthyroid cats in the analysis. The prevalence of hypercoagulability (as previously defined) was 14/23 (60.9%) in hyperthyroid cats and 4/13 (30.8%) in the control group, although no difference was found in the proportions of hyperthyroid and euthyroid cats that were hypercoagulable (P = .08). Also, the increase in serum T4 concentration was not a predictor of hypercoagulability (P = .53). In the 7 cats that were reassessed after RIT, 5 of those cats on initial evaluation before RIT were classified as hypercoagulable. Hypercoagulability resolved after RIT in 1 of the 5 cats when reevaluated.
Four of these 5 cats had abnormal echocardiograms on initial presen- the thrombus, but prethrombotic endothelial injury also could be a possibility. 8  Interestingly, an increase in AT activity was noted in hyperthyroid cats. This finding is in contrast to what would be expected because AT deficiency could lead to a hypercoagulable state. Increased AT activity has been seen in cats with acquired heart disease when compared to healthy controls and this increased activity has been attributed to AT behaving as an acute phase reactant. 18 Along with AT, fibrinogen behaves as a positive acute phase protein in response to inflammatory cytokines (interleukin-1, interleukin-6, tumor necrosis factor alpha). In our study and in a previous study of hyperthyroid cats, plasma fibrinogen concentration was increased in agreement with findings in hyperthyroid humans. 3,4,14 In humans, vWF : Ag has been described as an acute phase reactant that strongly correlated with increases in serum C-reactive protein concentration and normalized over time. 19 Hyperthyroidism in cats has been described histologically as occurring most commonly as a result of follicular cell adenoma and multinodular adenomatous hyperplasia, which is similar to toxic nodular goiter described in people. 11 In contrast, autoimmune hyperthyroidism (Grave's disease) also is described in humans. 20 8 Given that hyperfibrinogenemia has been found in cats with cardiomyopathies and our findings were associated with cardiac abnormalities, hyperfibrinogenemia may be driven by cardiac changes.
Overall, altered hemostasis in hyperthyroid cats does not seem solely attributable to cardiac abnormalities.
A limitation of our study was the small number of cats presented for reevaluation. Alternative approaches in recognizing hypercoagulability, such as viscoelastic tests and thrombin generation, were not performed and is a limitation of the study. Another possible limitation of our study is the fact that 2 cats had subclinical hypothyroidism (increased serum TSH concentration and normal serum T4 concentration) post-RIT. Overt hypothyroidism can promote a hypocoagulable state in people. 27 Given that the effects of subclinical hypothyroidism on hemostasis are unknown and our main goals were to evaluate change in hemostasis after resolution of hyperthyroidism, we did not exclude cats with subclinical hypothyroidism.
Whether subclinical hypothyroidism affects hemostasis in cats requires further study. Also, it is possible that in the enrollment of hyperthyroid cats, comorbidities such as renal or hepatic disease may have been present and could have contributed to apparent hypercoagulability. Another limitation was the lack of standardization for sedation. The sedation protocol was standardized, but the decision of whether or not to use sedation was made based on patient temperament.
In this group of hyperthyroid cats presented for RIT, although