Elsevier

The Veterinary Journal

Volume 179, Issue 1, January 2009, Pages 121-129
The Veterinary Journal

Tissue factor activated thromboelastography correlates to clinical signs of bleeding in dogs

https://doi.org/10.1016/j.tvjl.2007.08.022Get rights and content

Abstract

The ability of a laboratory assay to correlate to clinical phenotype is crucial for the accurate diagnosis and monitoring of haemostasis and is therefore challenging with currently used routine haemostasis assays. Thromboelastography (TEG) is increasingly used to evaluate haemostasis in humans and may well be of value in the workup of dogs suspected of having a haemostatic disorder. This study was undertaken to evaluate prospectively how tissue factor (TF) activated TEG correlated to clinical signs of bleeding in dogs, compared to a routine coagulation profile. A prospective case-control study was performed over a 2 year period from 2004–2006. Eligible dogs were those where the primary clinician requested a coagulation profile to evaluate haemostasis. The dogs were simultaneously evaluated with a TF-activated TEG assay. Twenty-seven dogs, characterised as hypo-coagulable based on the TEG parameter G (<3.2K dyn/cm2), were included in the study as cases. Size matched control groups of TEG normo- (G = 3.2K–7.2K dyn/cm2) and hyper-coagulable (G > 7.2K dyn/cm2) dogs were selected retrospectively from the eligible dogs. For all dogs, clinical signs of bleeding were noted at time of analysis.

There were statistically significant differences between all TEG values of hypo- and normo- and hyper-coagulable dogs. Thromboelastography correctly identified dogs with clinical signs of bleeding with a positive predictive value (PPV) of 89% and a negative predictive value (NPV) of 98% based on G alone. In comparison, the coagulation profile had a PPV between 50–81% and a NPV between 92–93% for detection of bleeding, depending on the observer. In conclusion, a TF-activated TEG G value < 3.2K dyn/cm2 correctly identified dogs with clinical signs of bleeding with very high PPV and NPV, irrespective of observer. The findings strongly suggest that TF- activated TEG may be of value in the workup of dogs suspected of having a haemostatic disorder.

Introduction

The ability of a laboratory assay to reveal and correlate to clinical phenotype is crucial for rational haemostasis monitoring. The ideal haemostasis assay should therefore be able to identify both biochemical and cellular abnormalities in the haemostasis system and at the same time correlate to the clinical signs of the patient.

Haemostasis has traditionally been perceived as a pathway divided into primary and secondary haemostasis, with the latter further divided into intrinsic and extrinsic parts. Until recently, this model was also viewed clinically as a model of physiological haemostasis. This view has persisted due to the fact that plasma-based haemostasis assays, such as activated partial thromboplastin time (aPTT), prothrombin time (PT), fibrinogen and D-dimer, are often used, along with a platelet count, in the routine diagnosis and monitoring of both humans and dogs with haemostatic disorders as if they are predictive of bleeding. In reality, the results are often in conflict with clinical signs, especially in mild or chronic cases (Zwack and Derkay, 1997, Asaf et al., 2001, Chowdhury et al., 2004, Segal and Dzik, 2005, Collins et al., 2006).

In humans, it has been found that measurement of PT, aPTT or bleeding time as a general preoperative screening procedure is neither cost-effective nor efficient for the identification of patients with increased risk of bleeding and, furthermore, that a normal platelet count does not exclude a severe platelet function defect (Dempfle, 2005). Thus, plasma-based haemostasis assays may accurately and reliably provide a specific biochemical diagnosis (as they are designed to), but the results may not necessarily correlate to the clinical phenotype of the patient. One reason for this disagreement could be that these tests are performed on platelet poor plasma (PPP) and therefore only provide information of the initiation of clot formation in a plasma environment.

The introduction of the cell-based model of haemostasis, where coagulation is initiated through exposure of cell-based tissue factor (TF) VIIa, has made it evident that because haemostasis is influenced by numerous other pro- and anticoagulant factors other than those present in blood plasma alone, it is essential that the methods used for the diagnosis and monitoring of patients take this into account. In particular, TF expression in certain types of tissue and cellular components of the blood, such as activated platelets and leukocytes, supply a surface for initiation, amplification and propagation of clot formation and thus play a key role in haemostasis (Hemker et al., 1983, Rosing et al., 1985, McVey, 1999, Hoffman and Monroe, 2001, Butenas et al., 2001). Furthermore, these cellular and tissue components are themselves influenced by altered systemic inflammatory and immune responses during disease (Osterud and Flaegstad, 1983, Burstein et al., 1996, Franco et al., 2000, Levi et al., 2004). Thus, the cell-based model of haemostasis has increased the understanding of the complex biochemistry of haemostasis and forced a re-evaluation of the traditional view of the intrinsic and extrinsic pathways of coagulation.

With the knowledge that whole blood contains all the intravascular factors and cells participating in physiological and pathological haemostasis, incorporating TF and phospholipid-bearing cells, it is reasonable to assume that whole blood assays, including thromboelastography (TEG), may provide a more accurate reflection of haemostasis in vivo than the traditionally used plasma-based haemostasis assays. Recently, a TF-activated TEG assay was validated in dogs (Wiinberg et al., 2005, Wiinberg et al., 2007) and preliminary results show that TEG tracings obtained from human blood in different states of haemostasis (Fig. 1) were also likely to apply to animals (Tuman et al., 1989, Tuman et al., 1993, Spiess et al., 1989, Mallett and Cox, 1992, Hoffman et al., 1993, Dürr and Kraft, 1997, Otto et al., 2000, Ao et al., 2001, Jessen et al., 2006).

The objectives of the present study were (1) to assess the overall haemostatic state in dogs clinically suspected to have a haemostatic disorder using both a coagulation profile and TF-activated TEG, (2) to examine the relationship between the results of TEG and the coagulation profile, and (3) to assess the ability of TEG and a conventional coagulation profile to detect signs of bleeding at the time of analysis.

Section snippets

Animals

The project was approved by the Small Animal Ethics and Administrative Committee at the Department of Small Animal Clinical Sciences, Faculty of Life Sciences, University of Copenhagen, Denmark and by the IACUC committee at Cummings School of Veterinary Medicine at Tufts University, North Grafton, MA, USA.

The investigation was designed as a prospective case-control study with hypo-coagulable TEG as the outcome variable and clinical signs of bleeding as the examined attribute. The work was

Results

Twenty-four out of 27 (89%) dogs in the hypo-coagulable group (G < 3.2K) had systemic clinical signs of bleeding. Only 1/27 (4%) of the dogs in the “normal” range (G 3.2K–7.2K) had clinical signs of bleeding (post operative haemarthrosis) and 0/27 (0%) of the dogs in the hyper-coagulable group (G > 7.2K) had clinical signs of bleeding. Of these, 27/27 (100%) of the dogs in the hypo-coagulable, 22/27 (81%) of the dogs in the normo-coagulable and 19/27 (70%) of the dogs in the hyper-coagulable groups

Discussion

This study demonstrated that TF-activated TEG is able to correctly identify dogs with clinical signs of bleeding, with both a higher PPV and higher NPV than a conventional coagulation profile. The results further indicated that, with a higher NPV and (especially) a higher PPV to detect bleeding, TEG G is easier for the clinician to interpret than the combination of tests comprising the coagulation profile used in this study.

Disorders of haemostasis are frequently encountered in critically ill

Conclusions

The TF-activated TEG assay correctly identified dogs with clinical signs of bleeding and does so better than a conventional coagulation screen. The findings suggested that patients suspected of having a haemostatic disorder could benefit from assessment of their overall haemostatic state prior to therapy, and that TF-activated TEG may be used as a valuable tool. Future prospective studies should evaluate whether TF-TEG is able to predict the risk of bleeding in patients, such as those

Acknowledgements

The authors wish to thank Natacha Errebo for expert technical assistance and the participating staff and patients at the Department of Small Animal Clinical Sciences, University of Copenhagen, Denmark, as well as Fisher Medical A/S and Novo Nordisk A/S for supporting this study.

References (38)

  • V.A. Bowbrick et al.

    Influence of platelet count and activity on thromboelastography parameters

    Platelets

    (2003)
  • S.A. Burstein et al.

    Cytokine-induced alteration of platelet and hemostatic function

    Stem Cells

    (1996)
  • S. Butenas et al.

    Platelets and phospholipids in tissue factor-initiated thrombin generation

    Thrombosis and Haemostasis

    (2001)
  • P. Chowdhury et al.

    Efficacy of standard dose and 30 ml/kg fresh frozen plasma in correcting laboratory parameters of haemostasis in critically ill patients

    British Journal of Haematology

    (2004)
  • P.W. Collins et al.

    Global tests of haemostasis in critically ill patients with severe sepsis syndrome compared to controls

    British Journal of Haematology

    (2006)
  • C.E. Dempfle

    Perioperative coagulation diagnostics

    Der Anaesthesist

    (2005)
  • Dürr, U.M., Kraft, W., 1997 Klinische Labordiagnostik in der Tiermedisin. 4. Überarbeitete und erweiterte Auflage...
  • J.L. Francis et al.

    Hemostasis and malignancy

    Seminars in Thrombosis and Hemostasis

    (1998)
  • C.B. Grindem et al.

    Thrombocytopenia associated with neoplasia in dogs

    Journal of Veterinary Internal Medicine

    (1994)
  • View full text