• Bacterial sepsis;
  • cancer;
  • coagulation;
  • procarboxypeptidase U;
  • thrombomodulin;
  • thrombosis


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Background: In people, increased thrombin-activatable fibrinolysis inhibitor (TAFI) antigen has been associated with increased risk of thrombosis, and decreased TAFI may contribute to bleeding diathesis. TAFI activity in dogs has been described in experimental models, but not in dogs with spontaneous disease.

Objective: The aim of this study was to compare TAFI activity in healthy dogs with TAFI activity in dogs with spontaneous disease.

Methods: Plasma samples from 20 clinically healthy Beagles and from 35 dogs with various diseases were analyzed using a commercial chromogenic assay that measured TAFI activity relative to activity in standardized pooled human plasma.

Results: Median TAFI activity for the 20 Beagles was 46.1% (range 32.2–70.8%) compared with 62.6% (29.1–250%) for the 35 diseased dogs, and 14/35 (40%) had TAFI activities >the upper limit for controls. The highest individual activities (>225%) were in 3 dogs with malignant neoplasms and 1 dog with thrombocytopenia. For data grouped by diagnosis, median TAFI activity was 61.7% for benign neoplasia (n=5), 64.9% for malignant neoplasia (n=8), 75.5% for Angiostrongylus vasorum infection (n=4), 68.8% for bacterial sepsis (n=7), and 58.7% for miscellaneous diseases (n=11). Compared with TAFI activity in control dogs, median TAFI activity was significantly increased only in the group of dogs with bacterial sepsis.

Conclusion: Bacterial sepsis was associated with significantly increased TAFI activity, and individual dogs with increased TAFI activities were found in all disease groups. The role of TAFI in the pathogenesis of hemostatic disorders in dogs and its value as a prognostic indicator deserve further investigation.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Thrombin-activatable fibrinolysis inhibitor (TAFI), also known as procarboxypeptidase U, was recently identified as a potent downregulator of fibrinolysis.1,2 Antifibrinolytic action of TAFI is primarily mediated through removal of carboxy-terminal lysine and arginine from fibrin, thereby decreasing plasminogen activation and attenuating positive feedback expressed by lys-plasminogen formation during fibrinolysis. However, at high concentrations TAFI also directly inhibits plasmin.3 Large concentrations of thrombin, generated by thrombin-mediated activation of factor XI, are required to activate TAFI.4,5 Thrombomodulin markedly enhances this activation, indicating that the thrombin-thrombomodulin complex is the likely physiologic activator of TAFI.6 TAFI antigen and activity may be measured in plasma based on binding of anti-TAFI antibody or substrate cleavage, respectively.

In people increased TAFI antigen has been associated with increased risk of venous thrombosis and recurrent pulmonary thromboembolism7,8 and has been found in various diseases with increased thrombotic potential, such as cancer,9,10 inflammatory bowel disease,11 and polycystic ovary syndrome.12 At the other end of the scale, decreased TAFI activity has been implicated as a contributor of bleeding diathesis in people with acute promyelocytic leukemia13 and in people with factor XI deficiency.4,14 Decreased TAFI antigen and TAFI activity have been identified in people with sepsis and infectious disease, respectively.15,16 Furthermore, decreased TAFI antigen has been identified in some studies of disseminated intravascular coagulation in people.15 A recent study suggests that TAFI may also serve as a downregulator of inflammation.17

TAFI represents an essential link between coagulation and fibrinolytic cascades, and determination of TAFI activity will likely provide new insight into the dynamics and pathogenesis of hemostatic disorders. TAFI activity in dogs has been described only in experimental models of thrombosis,18,19 and to the authors' knowledge there have been no studies of TAFI activity in dogs with spontaneous disease. Thus, the aim of this exploratory study was to measure and compare TAFI activity in healthy dogs and in dogs with spontaneous diseases, including some in which hemostatic aberrations were likely to be present.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References


The study 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 the Cummings School of Veterinary Medicine at Tufts University, North Grafton, MA, USA. Fifty-five dogs were included in this study. The control group included 20 clinically healthy dogs from a research colony of beagle dogs. Criteria for determining that dogs were healthy were an unremarkable routine physical examination and no abnormalities detected on a CBC, biochemical profile, and coagulation profile consisting of prothrombin time (PT), activated partial thromboplastin time (aPTT), d-dimer assay, and fibrinogen concentration. The remaining 35 dogs were determined to have various clinical abnormalities. Eighteen dogs belonged to a population of dogs admitted to the intensive care unit (ICU) at Tufts University between May and July 2006. Seven of these 18 dogs were diagnosed with bacterial sepsis and had clinical signs of systemic inflammatory response syndrome in combination with identification of a pathogen either by means of a culture from blood or abdominal effusion or by means of a cytologic identification of intracellular bacteria in abdominal effusion. A diagnosis of systemic inflammatory response syndrome was made when 2 or more of the following criteria were fulfilled: hyperthermia or hypothermia, tachypnea, tachycardia, and neutropenia or neutrophilia. Thirteen dogs were diagnosed with neoplasia at the University of Copenhagen, Denmark, between November and December 2005. Neoplasia was confirmed by histopathologic examination, and the tumors were subdivided into into benign and malignant processes. Four dogs belonged to a group of dogs diagnosed with Angiostrongylus vasorum infection (French Heartworm) at the University of Copenhagen between April 2005 and September 2006. A. vasorum infection was diagnosed by identification of larvae in feces by means of the Baermann flotation test.

Blood collection and plasma storage

Blood was collected into 2.8 mL plastic tubes containing 3.2% sodium citrate (Vacuette, Greiner Bio-One International AG, Kremsmunster, Austria), and plasma was separated by centrifugation at 4000g for 120 seconds and aliquoted into cryovials for storage. Vials containing plasma from the18 dogs in the ICU at Tufts University were transported to Copenhagen, Denmark on dry ice by certified courier. Plasma from 20 control dogs was sampled 2 months before analysis. Plasma from the other dogs was, with the exception of plasma from A. vasorum dogs, sampled between 19 and 21 months before analysis. For the 4 dogs with A. vasorum infection, plasma was collected 14–32 months before analysis. All plasma samples were stored at −80°C and thawed once before analysis. Samples from dogs with neoplastic disease were assayed separately along with 6 internal controls. All other samples were assayed in the same run.

TAFI assay

Analysis of canine TAFI activity was performed with a commercially available chromogenic assay (Pefakit TAFI, Pentafarm, Basel, Switzerland). The assay is designed as a kinetic enzymatic assay assessing enzymatic activity of TAFI activated by thrombin-thrombomodulin calibrated with pooled plasma from healthy human donors (standardized to the Secondary Coagulation Standard [SSC/ISTH lot 2]). The functional assay is based on direct cleavage of a peptide mimetic substrate. The assay was performed according to manufacturer's instructions. Calibrators and controls provided by the manufacturer were reconstituted in 1.0 mL of deionized water and incubated in closed vials for 10 minutes at 21°C (room temperature). Frozen plasma samples were thawed for 10 minutes at 37°C and subsequently diluted 1:2 with deionized water. Ten μL of diluted plasma were mixed with 100 μL of reagent 1 (thrombin-thrombomodulin) and incubated for 3 minutes at 37°C. At the end of incubation 100 μL of reagent 2 (synthetic substrate) was added to the solution followed by immediate measurement of absorbance at 405 nm every 10 seconds for 5 minutes in an automated plate reader (Multiscan RC, Thermo Scientific, Waltham, MA, USA). All measurements were run in duplicates. TAFI activity was expressed as percent relative to activity of standardized pooled human plasma provided as a calibrator.

TAFI assay characteristics

To assess assay imprecision at relevant activity levels 3 pools were established by mixing plasma from 3 to 4 dogs at each level. Sixteen determinations of each pool were performed in a single run to assess intra-assay imprecision. Interassay imprecision was assessed by performing duplicate measurements of each pool in 5–8 separate runs. Only vials needed for each specific analytical run were thawed to prevent potential variation as a result of repetitive freeze–thaw cycles. Repetitive measurements of blanks (n=16) across several runs were performed to assess the detection limit.

Statistical analysis

Medians, means, intra- and interassay variations, and coefficients of variation (CV) were calculated using routine descriptive statistical procedures. Detection limit was assessed as mean activity of blanks plus 3 SD. Mann–Whitney analysis was applied for the purpose of comparing TAFI activity in different groups of diseased dogs with that of healthy dogs. Statistical significance was set at P<.05. A computer software program (GRAPHPAD PRISM v. 4.01, GraphPad Software, San Diego, CA, USA) was used for statistical analysis.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The observed intra- and interassay imprecision at various TAFI activities is given in Table 1. Detection limit of the assay was assessed to 7.4% (mean 2.6%, SD 1.6%). Data from dogs were retrospectively assigned to 5 different groups based on diagnoses: benign neoplasia (5 dogs), malignant neoplasia (8 dogs), angiostrongylosis (4 dogs), bacterial sepsis (7 dogs), and miscellaneous diseases (11 dogs). Sex, breed, age, final diagnosis, and TAFI activity of all dogs are listed in Table 2.

Table 1.   Intra- and interassay imprecision at various TAFI activities in canine plasma samples.
 No. of SamplesMean Value (%)SD (%)CV (%)
  1. CV, coefficient of variation.

Low pool1634.52.26.2
Medium pool1663.13.04.7
High pool16159.412.47.8
Low pool741.97.417.6
Medium pool779.110.012.6
High pool7186.022.212.0
Table 2.   Signalment, diagnosis, and thrombin-activatable fibrinolysis inhibitor (TAFI) activity of healthy Beagles and 35 dogs with diseases.
 BreedAgeSexClinical StatusTAFI Activity (%)
  1. Bolded TAFI activities are above the upper limit for the healthy Beagles.

  2. F, female; M, male; MC, castrated male; FS, spayed female; GDV, gastric dilatation-volvulus.

Controls (n=20)Beagles2–5 years12 F 8 MCClinically healthy32.2–70.9 (range)
Malignant neoplasia
 1Dachshund1.5 yearsFLymphoma (stage IVa)250.0
 2Bullmastiff9 yearsMCLymphoma (stage IIIb)60.5
 3Dachshund9 yearsFMammary adenocarcinoma (stage II)35.1
 4West Highland White Terrier13 yearsFBronchoalveolar carcinoma227.0
 5American Bulldog4.5 yearsMCMalignant mesenchymal tumor (oral)58.0
 6Labrador Retriever9 yearsFNasal carcinoma38.0
 7Mixed breed8 yearsMScrotal mastocytoma grade III (stage I)250.0
 8Cairn Terrier4 yearsMCCutaneous melanoma69.2
Benign neoplasia
 9Maltese8.5 yearsFSMammary gland adenoma55.2
10Golden Retriever9.5 yearsFMammary gland adenoma88.3
11Rhodesian Ridgeback8.5 yearsFMammary gland adenoma and lipoma81.7
12Border Collie2.5 yearsFSSquamous papilloma (vulva)61.7
13Mixed breed10 yearsFSSkin hamartoma, not infected or inflamed37.7
Infectious diseases
14Basset Fauve de Bretagne1.3 yearsMAngiostrongylus vasorum54.9
15Labrador Retriever1 yearMAngiostrongylus vasorum39.6
16Dachshund14 yearsFSAngiostrongylus vasorum117.0
17Mixed breed6 monthsFAngiostrongylus vasorum96.1
18Portuguese Water Dog12 yearsFSBacterial sepsis82.9
19Dachshund14.2 yearsFSBacterial sepsis110.0
20Basset Hound9 yearsMBacterial sepsis29.1
21Dalmatian9.8 yearsFSBacterial sepsis57.7
22Doberman10 yearsFSBacterial sepsis62.6
23Mixed breed12 yearsFSBacterial sepsis68.8
24Australian Shepherd8 yearsMCBacterial sepsis96.1
Miscellaneous diseases
25Standard Poodle6.6 yearsFSGDV40.5
26German Shepherd1.7 yearsMCHemorrhagic gastroenteritis29.1
27Mixed breed6 yearsMHeatstroke140.0
28Dachshund12.8 yearsMBite wound70.8
29Labrador Retriever12 yearsMPancreatitis119.0
30West Highland White Terrier5 yearsFSSnake bite30.6
31Border Collie9.3 yearsMCThrombocytopenia58.7
32Labrador Retriever9 yearsMCThrombocytopenia250.0
33Labrador Retriever9.8 yearsFSHit by car32.2
34Labrador Retriever6.9 yearsMCGDV with gastric perforation98.6
35Boxer1.2 yearsFSHit by car40.5

Median TAFI activity in control dogs was 46.1% (range 32.2–70.8%). Median (range) for all dogs diagnosed with clinical abnormalities was 62.6% (29.1–250%). TAFI activity in dogs with clinical abnormalities was significantly increased compared with control dogs (P=0.023). There were 14 of 35 (40%) affected dogs with values above the range of controls and 3 dogs (8.6%) with values slightly below. The highest individual TAFI activities (above 225%) were observed in 3 dogs with malignant neoplasms and 1 dog with thrombocytopenia. Analyzed by groups, median (range) of TAFI activities were: 61.7% (37.7–88.3%) for benign neoplasia; 64.9% (35.1–250%); for malignant neoplasia; 75.5% (39.6–117%) for A. vasorum infection; 68.8% (29.1–110%) for bacterial sepsis; and 58.7% (29.1–250%) for miscellaneous diseases (Figure 1). Compared with TAFI activity in control dogs, median TAFI activity was significantly increased only in the group of dogs with bacterial sepsis (P=.025; Figure 1), but only 3 of those 7 dogs had values greater than the upper limit of controls and 1 dog had a low value.


Figure 1.  Box and whisker plot of thrombin-activatable fibrinolysis inhibitor (TAFI) activity in healthy dogs and in dogs with various diseases. The top and bottom of the box indicate the upper and lower quartile, respectively, and the horizontal line is the median value. Whiskers are minimum and maximum values. Compared with healthy dogs, P-values are .023 for all diseases, .126 for benign neoplasia, .088 for malignant neoplasia, .188 for angiostrongylosis, .025 for bacterial sepsis, and .563 for miscellaneous diseases.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

More than one-third of the dogs in this study had TAFI activities above the upper limit of control dogs. Interestingly, 3 of the 4 TAFI activities of greatest magnitude were found in dogs with malignant neoplasms. In people significantly increased TAFI activities and plasma antigen concentrations have been found to contribute to the hypercoagulable state in various types of lung cancers.9,10 A recent study of 30 dogs with neoplasia demonstrated that the majority of dogs with malignant neoplasms are hypercoagulable.20 Whether the same relationship between malignancies, thrombosis, and TAFI activity exists in dogs has yet to be investigated.

Observed TAFI activities in dogs with bacterial sepsis in our study were significantly different from TAFI activities in healthy dogs. Knowledge of the role of TAFI in host responses to severe infections is limited. In people reduced levels of TAFI antigen and TAFI activity have been identified in septic patients and in patients with infectious disease, respectively.15,16 In the sepsis study a low median TAFI antigen level was identified in a group of septic patients compared with a group of nonseptic patients though most of the individual TAFI antigen levels in septic patients were within the reference interval.16 In that study low TAFI antigen levels were speculated to be secondary to consumption due to a positive correlation between TAFI antigen and various coagulation factors (FII and FXI) though there was no correlation between overt DIC and TAFI antigen levels.16 Other possible explanations for low TAFI antigen levels in human sepsis include extravasation of proteins due to increased vascular permeability and genetic polymorphisms. In contrast upregulation of TAFI mRNA and circulating TAFI plasma concentrations have been found in response to endotoxin exposure in rodents and in a murine model of Escherichia coli peritonitis.21 In the latter model TAFI plasma antigen levels decreased by 20% in the early phase of infection but increased 2.5-fold after 20 hours.21 One out of 7 septic dogs in our study had a TAFI activity below the minimum value for healthy dogs and 3/7 septic dogs had TAFI activities above the maximum level of healthy dogs. Based on these results it seems reasonable to assume that bacterial sepsis in dogs likely leads to upregulation of TAFI at some stage and that plasma TAFI activity is determined by the balance between production and consumption or loss at a given point in time.

In our study a wide range of TAFI activity was established in a homogenous group of healthy research beagle dogs. Polymorphism in the promoter gene of TAFI has been identified in people; however, variation in TAFI antigen levels in healthy people22 may correlate with immunoreactivity of various isoforms when measuring antigen concentrations rather than being true differences in activity.23 Other risks of assay-related bias have been identified, but the application of an activity-based kinetic assay with substrates selectively detecting TAFI activity in the present study minimized those risks.23 Whether breed-related differences in TAFI antigen levels or TAFI activity occur in dogs is currently unknown. However, use of an unmatched control group with regard to phenotype, genotype, and environment is a weakness of this study. For future clinical studies, TAFI reference intervals should be established in a larger group of dogs that represent the population of dogs where the assay is applied.

The assay used in this study measured TAFI with intra-assay variation that is comparable to that reported by the manufacturer for human plasma samples and is considered acceptable for diagnostic purposes. On the other hand imprecision, as indicated by the interassay variation, may pose a problem if the assay is used for diagnostic purposes. The assay will only be able to consistently detect differences exceeding the interassay CV, and there is substantial overlap of TAFI activities between groups of possible clinical interest. However, running samples in batches permits detection of much smaller differences between groups. Thus, the assay seems applicable for group comparison of TAFI activity in future studies of the pathophysiologic significance of abnormal TAFI activity. The detection limit appears acceptable as it was well below observed activities. A study of linearity was not performed in the present study to confirm the quantitative nature of the assay. However, it is assumed that observed levels in the study are ranked correctly, and thus assumptions of nonparametric statistics are fulfilled.

In this study we used plasma that had been stored for up to 21 months for the majority of the diseased dogs and 32 months for a single dog with A. vasorum infection. The effect of long-term plasma storage on TAFI activity is unknown and could potentially result in protein degradation. Thus, prolonged storage and differences in plasma storage time between control and disease groups are weaknesses of this study. However, except for the dogs with A. vasorum infection, the disease groups were comparable with regard to plasma storage time both within and between groups, and the effect of prolonged storage should theoretically affect all groups in a similar way. If the effect of storage time on TAFI activity was substantial enough to cause a statistically significant difference between control and disease groups, we would have expected to see this difference in all disease groups, not the sepsis group only. Therefore, the significant difference between the sepsis and controls was attributed to factors associated with sepsis.

In conclusion, bacterial sepsis was associated with significantly increased TAFI activity, and at the same time individual dogs with abnormal TAFI activities were found in all disease groups. Further investigation of TAFI activity may help elucidate the complex pathophysiologic mechanisms leading to hemostatic dysfunction in disease. The role of TAFI in the pathogenesis of thrombosis and hemorrhagic diathesis in dogs with primary hemostatic disorders and hemostatic disorders secondary to systemic, inflammatory, or neoplastic diseases deserves further investigation. The clinical consequences of increased or decreased TAFI activity is currently unknown for dogs, and changes in activity potentially have prognostic implications for individual animals.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Disclosure: The authors have indicated that they have no affiliations or financial involvement with any organization or entity with a financial interest in, or in financial competition with, the subject matter or materials discussed in this article.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Eaton DL, Malloy BE, Tsai SP, Henzel W, Drayna D. Isolation, molecular cloning, and partial characterization of a novel carboxypeptidase B from human plasma. J Biol Chem. 1991;266:2183321838.
  • 2
    Wang W, Hendriks DF, Scharpe SS. Carboxypeptidase U, a plasma carboxypeptidase with high affinity for plasminogen. J Biol Chem. 1994;269:1593715944.
  • 3
    Wang W, Boffa MB, Bajzar L, Walker JB, Nesheim ME. A study of the mechanism of inhibition of fibrinolysis by activated thrombin-activable fibrinolysis inhibitor. J Biol Chem. 1998;273:2717627181.
  • 4
    Bouma BN, Marx PF, Mosnier LO, Meijers JC. Thrombin-activatable fibrinolysis inhibitor (TAFI, plasma procarboxypeptidase B, procarboxypeptidase R, procarboxypeptidase U). Thromb Res. 2001;101:329354.
  • 5
    Von dem Borne PA, Bajzar L, Meijers JC, Nesheim ME, Bouma BN. Thrombin-mediated activation of factor XI results in a thrombin-activatable fibrinolysis inhibitor-dependent inhibition of fibrinolysis. J Clin.Invest. 1997;99:23232327.
  • 6
    Bajzar L, Morser J, Nesheim M. TAFI, or plasma procarboxypeptidase B, couples the coagulation and fibrinolytic cascades through the thrombin-thrombomodulin complex. J Biol Chem. 1996;271:1660316608.
  • 7
    Van Tilburg NH, Rosendaal FR, Bertina RM. Thrombin activatable fibrinolysis inhibitor and the risk for deep vein thrombosis. Blood. 2000;95:28552859.
  • 8
    Eichinger S, Schonauer V, Weltermann A, et al. Thrombin-activatable fibrinolysis inhibitor and the risk for recurrent venous thromboembolism. Blood. 2004;103:37733776.
  • 9
    Hataji O, Taguchi O, Gabazza EC, et al. Increased circulating levels of thrombin-activatable fibrinolysis inhibitor in lung cancer patients. Am J Hematol. 2004;76:214219.
  • 10
    Koldas M, Gummus M, Seker M, et al. Thrombin-activatable fibrinolysis inhibitor levels in patients with non-small-cell lung cancer. Clin Lung Cancer. 2008;9:112115.
  • 11
    Saibeni S, Bottasso B, Spina L, et al. Assessment of thrombin-activatable fibrinolysis inhibitor (TAFI) plasma levels in inflammatory bowel diseases. Am J Gastroenterol. 2004;99:19661970.
    Direct Link:
  • 12
    Karakurt F, Gumus II, Bavbek N, et al. Increased thrombin-activatable fibrinolysis inhibitor antigen levels as a clue for prothrombotic state in polycystic ovary syndrome. Gynecol Endocrinol. 2008;24:491497.
  • 13
    Meijers JC, Oudijk EJ, Mosnier LO, et al. Reduced activity of TAFI (thrombin-activatable fibrinolysis inhibitor) in acute promyelocytic leukaemia. Br J Haematol. 2000;108:518523.
  • 14
    Bouma BN, Mosnier LO. Thrombin activatable fibrinolysis inhibitor (TAFI) at the interface between coagulation and fibrinolysis. Pathophysiol Haemost Thromb. 2003;33:375381.
  • 15
    Watanabe R, Wada H, Watanabe Y, et al. Activity and antigen levels of thrombin-activatable fibrinolysis inhibitor in plasma of patients with disseminated intravascular coagulation. Thromb Res. 2001;104:16.
  • 16
    Zeerleder S, Schroeder V, Hack CE, Kohler HP, Wuillemin WA. TAFI and PAI-1 levels in human sepsis. Thromb Res. 2006;118:205212.
  • 17
    Myles T, Nishimura T, Yun TH, et al. Thrombin activatable fibrinolysis inhibitor, a potential regulator of vascular inflammation. J Biol Chem. 2003;278:5105951067.
  • 18
    Bjorkman JA, Abrahamsson TI, Nerme VK, Mattsson CJ. Inhibition of carboxypeptidase U (TAFIa) activity improves rt-PA induced thrombolysis in a dog model of coronary artery thrombosis. Thromb Res. 2005;116:519524.
  • 19
    Wang YX, Da Cunha V, Vincelette J, et al. A novel inhibitor of activated thrombin activatable fibrinolysis inhibitor (TAFIa)—part II: enhancement of both exogenous and endogenous fibrinolysis in animal models of thrombosis. Thromb Haemost. 2007;97:5461.
  • 20
    Kristensen AT, Wiinberg B, Jessen LR, Andreasen E, Jensen AL. Evaluation of human recombinant tissue factor-activated thromboelastography in 49 dogs with neoplasia. J Vet Intern Med. 2008;22:140147.
  • 21
    Renckens R, Roelofs JJ, Ter Horst SA, et al. Absence of thrombin-activatable fibrinolysis inhibitor protects against sepsis-induced liver injury in mice. J Immunol. 2005;175:67646771.
  • 22
    Chetaille P, Alessi MC, Kouassi D, Morange PE, Juhan-Vague I. Plasma TAFI antigen variations in healthy subjects. Thromb Haemost. 2000;83:902905.
  • 23
    Willemse JL, Hendriks DF. Measurement of procarboxypeptidase U (TAFI) in human plasma: a laboratory challenge. Clin Chem. 2006;52:3036.