SEARCH

SEARCH BY CITATION

Standardizing the risk of bleeding remains an important clinical goal, and ideally should take into account coagulation factors, platelet functionality, and vascular integrity issues. One such test to which a great deal of energy has been devoted for standardization has been the human template bleeding time [1]. This test now routinely involves using a precise forearm incision under specific blood pressure conditions, but in spite of such measures, this test is not as good as a careful history in predicting risk of bleeding with surgery [2,3]. When murine models of hemostatic and thrombotic disorders were developed, it was hoped that, because murine strain as well as the underlying hemostatic defect could be carefully controlled, better prediction would be obtained. The model most often selected involved a tail injury [4,5]. This was an attractive model, as researchers were already using tail snips for genomic identification. However, the details of this model have never been well standardized. On the basis of our own experience, discussions with other investigators in the field of murine hemostasis, and the available literature, we believe that important variables include the following.

Strain, sex and age variables

  1. Top of page
  2. Strain, sex and age variables
  3. Anesthetic variables
  4. Tail injury variables
  5. Outcome variables
  6. Published utility in hemostatic models
  7. Disclosure of Conflict of Interests
  8. References

Different strains of mice vary greatly in platelet counts and in other coagulation parameters. It is not surprising that the bleeding time varies between strains. One study found that the average bleeding time was 51 s in C57Bl/6 mice and 62 s in 129/Sv mice [6]. Moreover, the two strains showed a marked difference in the prolongation of the bleeding time when treated with aspirin or heparin. The age and sex of mice can affect the bleeding time as well. Male mice (and their tails) tend to be larger than their female counterparts at any given age. Using mice within a certain body mass range instead of mice of a certain age would help to ensure that these variables are kept as constant as possible.

Anesthetic variables

  1. Top of page
  2. Strain, sex and age variables
  3. Anesthetic variables
  4. Tail injury variables
  5. Outcome variables
  6. Published utility in hemostatic models
  7. Disclosure of Conflict of Interests
  8. References

Anesthesia varies between groups: the use of isoflurane [7], ketamine/medetomidine, ketamine/xylazine or pentobarbiturate [8] has been described [6]. The influences of such different anesthetics on bleeding need to be evaluated, as these different anesthetics affect systolic blood pressure differently [9], and differences in blood pressure could influence blood loss.

Body temperature changes are another major concern. Anesthetized mice rapidly cool, affecting systemic blood flow distribution [9,10]. It is our impression that most groups place the mice on a heating pad to prevent hypothermia, but as only the ventral body is warmed, mice still tend to become hypothermic. This may lead to peripheral vasoconstriction. Placing the mice in an evenly warmed microenvironment, such as a warming chamber, may be an improvement in preventing hypothermia, and also warrants further analysis.

Tail injury variables

  1. Top of page
  2. Strain, sex and age variables
  3. Anesthetic variables
  4. Tail injury variables
  5. Outcome variables
  6. Published utility in hemostatic models
  7. Disclosure of Conflict of Interests
  8. References

Selection of where to make the incision site also varies between groups. Some laboratories incise the tail at a fixed distance from the tip [8], whereas others incise the tail at a specific diameter [6]. Some groups only incise the tail veins rather than cut off the entire tail tip. The sharpness of the scalpel and/or surgical knife are likely to vary as well. Reused scalpels and surgical scissors may well introduce an element of crush injury as well as laceration. One way to avoid a dull blade effect is to use a tail-guillotine, with which the tail is cut at a fixed distance from the tip with a fixed downwards force [11]. The blade is exchanged after each treatment group to avoid crush injury.

Collecting the blood into warmed saline is currently the most accepted technique, but dabbing with a Whatman paper until bleeding stops has also been used, as well as just observing the dependent tail. This tail dabbing may introduce bias, as people may dab and traumatize the tail of a particular subgroup more than that of another, and induce longer bleeding. Ideally, mice should be studied with the operator blinded to the experimental condition of the mouse.

Outcome variables

  1. Top of page
  2. Strain, sex and age variables
  3. Anesthetic variables
  4. Tail injury variables
  5. Outcome variables
  6. Published utility in hemostatic models
  7. Disclosure of Conflict of Interests
  8. References

Blood can be gravimetrically collected and observed to measure both the volume and time to cessation of bleeding [7]. Alternatively blood loss as measured by hemoglobin concentration with the tail dipped into warmed saline [8] can be measured as can the ability to survive the tail bleed [12]. Bleeding time may be defined as the time point of first cessation of bleeding, but rebleeds often occur within a few minutes, in many cases alternating several times with short cessation of bleeding. In assessment of total blood loss, the observation time may vary greatly between the laboratories.

Published utility in hemostatic models

  1. Top of page
  2. Strain, sex and age variables
  3. Anesthetic variables
  4. Tail injury variables
  5. Outcome variables
  6. Published utility in hemostatic models
  7. Disclosure of Conflict of Interests
  8. References

Our impression and that of one publication is that the bleeding time described above tends to be normal in a surprisingly wide variety of coagulation defects, although it tends to be prolonged in severe platelet disorders [6]. The reason for this is unclear, but bleeding may stop owing to vessel spasm and the platelet plug being able to temporarily control bleeding in this low-pressure, low-flow state. Clearly, mice with coagulation disorders tend to have late and repeated bleeding, especially after the animal has recovered from surgery. These mice naturally lick and handle their injured tail, which probably dislodges unstable clots, especially as vessel spasm coincidentally recedes. Factor VIII null mice and other hemophilic mice exsanguinate by the next day if they do not have their cut tails cauterized [12].

For several groups, the ability to prevent mice with severe hemophilia from bleeding to death after tail vein resection has become the standard for demonstrating effective non-conventional therapy for hemophilia [13,14], but this should be viewed with caution, because several hours after a similar cuticular injury [15], hemophilic mice lose 30–70% of their blood volume, and we suspect that the same is true after a tail vein bleed. These animals are probably undergoing hypovolemic shock, and are shunting blood away from their tail, allowing otherwise minimal therapies to improve survival. Considering the need to incorporate humane endpoints into animal studies, for animal welfare reasons, ‘mortality’ as an endpoint has also to be questioned from an ethical point of view [16,17]. As there is a correlation between amount of total blood loss over time and mortality, alternative endpoints should be considered. The amount of blood loss within a time period of 6–8 h after injury could be taken as a surrogate for survival.

In summary, tail vein bleeding in mice is presently the most commonly used bleeding challenge for assessment of hemostatic state. Its biggest challenges are the need to standardize multiple aspects of the study mice and their handling during and after the procedure. Studies of littermates, especially prior to molecular discrimination of affected from non-affected siblings, are likely to be most meaningful. Because there are so many subtle variables regarding where and how the injury is inflicted and how the blood is collected, it is hard to compare results between laboratories at present. An attempt to standardize the technical setup of the tail bleeding model would first require interlaboratory tests using the same mouse strain and testing procedures. On the basis of these results, the influence of the different setups could be evaluated, and thus provide a basis for scientific discussion on how to standardize the model.

Even if standardized, the test may, in the end, be unable to detect very important bleeding conditions, because unstable clots in the presence of vascular spasm may be sufficient to temporarily protect against bleeding. These observations should lead to caution in drawing broad conclusions about the bleeding diathesis of a particular modification in mice based on one model under one set of challenges. Often, complementary models and/or appropriately performed and interpreted modifications are needed for full appreciation of the importance of a particular hemostasis/thrombosis model. As the tail bleeding time requires no sophisticated equipment and is one of the most widely used models, standardization is important to allow better interlaboratory communication and understanding of whether a particular genoytpe does or does not have a significant bleeding diathesis. At the same time, investigators are urged to appreciate the strengths and limitations of each hemostatic assay, and not limit themselves to a single assay.

References

  1. Top of page
  2. Strain, sex and age variables
  3. Anesthetic variables
  4. Tail injury variables
  5. Outcome variables
  6. Published utility in hemostatic models
  7. Disclosure of Conflict of Interests
  8. References
  • 1
    Nilsson IM, Magnusson S, Borchgrevink C. The Duke and Ivy methods for determination of the bleeding time. Thromb Diath Haemorrh 1963; 10: 22334.
  • 2
    Lind SE. The bleeding time does not predict surgical bleeding. Blood 1991; 77: 254752.
  • 3
    Watson HG, Greaves M. Can we predict bleeding? Semin Thromb Hemost 2008; 34: 97103.
  • 4
    Dejana E, Callioni A, Quintana A, De Gaetano G. Bleeding time in laboratory animals. II – A comparison of different assay conditions in rats. Thromb Res 1979; 15: 1917.
  • 5
    Novak EK, Sweet HO, Prochazka M, Parentis M, Soble R, Reddington M, Cairo A, Swank RT. Cocoa: a new mouse model for platelet storage pool deficiency. Br J Haematol 1988; 69: 3718.
  • 6
    Broze GJ Jr, Yin ZF, Lasky N. A tail vein bleeding time model and delayed bleeding in hemophiliac mice. Thromb Haemost 2001; 85: 7478.
  • 7
    Landskroner KA, Cui ZH, Newgren J, Fournel M, Pierce GF, Jesmok GJ. Evaluation of PEG-FVIII molecules with prolonged half-lives in a murine FVIII-dependent bleeding model. Blood 2006; 108, 464a.
  • 8
    Tranholm M, Kristensen K, Kristensen AT, Pyke C, Rojkaer R, Persson E. Improved hemostasis with superactive analogs of factor VIIa in a mouse model of hemophilia A. Blood 2003; 102: 361520.
  • 9
    Flecknell P. Laboratory Animal Anaesthesia, 2nd edn. Burlington, MA: Elsevier Academic Press, 1996.
  • 10
    Taylor DK. Study of two devices used to maintain normothermia in rats and mice during general anesthesia. JAALAS 2007; 46: 3741.
  • 11
    Leidenmuehler P, Resch M, Bischetsrieder B, Schiviz A, Wuersch K, Hoellriegl W, Schwarz HP, Muchitsch EM. Feasibility of standardized cut methods in the murine tail-clip bleeding assay – comparison of scalpel, tail guillotine and a contactless laser system. Poster presented at the 47th annual meeting of the Society for Laboratory Animal Science GV-SOLAS, 2009.
  • 12
    Bi L, Lawler AM, Antonarakis SE, High KA, Gearhart JD, Kazazian HH Jr. Targeted disruption of the mouse factor VIII gene produces a model of haemophilia A. Nat Genet 1995; 10: 11921.
  • 13
    Shi Q, Wilcox DA, Fahs SA, Weiler H, Wells CW, Cooley BC, Desai D, Morateck PA, Gorski J, Montgomery RR. Factor VIII ectopically targeted to platelets is therapeutic in hemophilia A with high-titer inhibitory antibodies. J Clin Invest 2006; 116: 197482.
  • 14
    Ohmori T, Ishiwata A, Kashiwakura Y, Madoiwa S, Mitomo K, Suzuki H, Hasegawa M, Mimuro J, Sakata Y. Phenotypic correction of hemophilia A by ectopic expression of activated factor VII in platelets. Mol Ther 2008; 16: 135965.
  • 15
    Yarovoi H, Kufrin D, Eslin DE, Zhu H, Wilcox DA, Camire R, Fakharzadeh SS, Kowalska MA, Montgomery RR, Poncz M. Factor VIII ectopically expressed in platelets. Efficacy in hemophilia A treatment. Blood 2003; 102: 400613. .
  • 16
    Organization for Economic Co-operation and Development Guidance Document on the Recognition, Assessment, and Use of Clinical Signs as Humane Endpoints for Experimental Animals Used in Safety Evaluation. OECD Environmental Health and Safety Publications Series on Testing and Assessment No. 19 ENV/JM/MONO(2000)7.
  • 17
    Wallace J. Humane endpoints in cancer research. ILAR J 2000; 41: 8793.