Disseminated intravascular coagulation;
Development and standardizationof a non-clinical rabbit model
Ph.D. Thesis
Line Olrik Berthelsen, DVM
Department of Small Animal Clinical Sciences
Faculty of Life Sciences
CopenhagenUniversity
Haemostasis Pharmacology
Novo Nordisk A/S
Copenhagen, Denmark
2010
TABLE OF CONTENTS:
PREFACE.……………………………………………………………………………………………………………………..3
ABBREVIATIONS…………………………………………………………………………………………………………….4
SUMMARY (ENGLISH)………………………………………………………………………………………………………5
SAMMENDRAG (DANSK)…………………………………………………………………………………………………...7
1Introduction, hypotheses and objectives
2Basic Mechanisms of Haemostasis and Thrombosis
2.1Physiologic Haemostasis
2.2Regulation of Coagulation
2.2.1Anticoagulation
2.2.2Fibrinolysis
2.2.2.1Activation of plasmin
2.2.2.2Regulation of the fibrinolytic system
2.3Acquired Procoagulant Disorders of Haemostasis
2.4Disseminated Intravascular Coagulation
2.4.1Aetiology
2.4.2Pathophysiology
2.4.3Interaction between inflammation and haemostasis
2.4.4Diagnosis of DIC
2.4.4.1Detection of Activation of Coagulation
2.4.4.2Inhibitor consumption
2.4.4.3Fibrinolytic activity
2.4.4.4ISTH classification of overt and non-overt human DIC
3Experimental thrombosis and DIC in animal models
3.1Methods in diagnosis of experimental microthrombosis
3.2Standardization and translational aspects of DIC in animal models
3.3Animal models of DIC and their relevance to human DIC – A systematic review ...... (Paper I)
4Development of a rabbit model of DIC and implementation of new parameters for early diagnosis of microthrombosis in DIC
4.1Validation of a pulmonary function test in a rabbit model of embolisation mimicked by microspheres
4.1.1Background
4.1.1.1Multiple Inert Gas Elimination Technique (MIGET)
4.1.2Materials and methods
4.1.2.1Animal procedures
4.1.2.2Microspheres
4.1.2.3Ventilation-perfusion relationships
4.1.2.4Experimental protocol
4.1.2.5Preparation of tissue specimens
4.1.3Results
4.1.3.1Histological examination
4.1.4Discussion and conclusion
4.1.5Cardiovascular and haemostatic changes in a rabbit microsphere model of pulmonary thrombosis (Paper II)
4.2Implementation of new parameters in early diagnosis of DIC
4.2.1Background
4.2.2Materials and methods
4.2.2.1Animals
4.2.2.2Animal procedures
4.2.2.3Cardiac troponin I (cTnI) measurements
4.2.2.4Thromboelastography (TEG)
4.2.2.5Experimental design
4.2.3Results
4.2.4Discussion and conclusion
4.3Establishment of a rabbit model of thromboplastin induced DIC
4.3.1Background
4.3.2Development of a model of thromboplastin induced DIC in rabbits
4.3.2.1Background
4.3.2.2Materials and methods
4.3.2.2.1Experimental design
4.3.2.3Results
4.3.2.4Conclusion
5Implementation of the ISTH classification of non-overt DIC in a thromboplastin induced rabbit model (Paper III)
6Characterisation and purification of thromboplastin
6.1Background
6.2Materials and methods
6.3Results
6.4Conclusion
7Purified thromboplastin causes haemostatic abnormalities but not overt DIC in an experimental rabbit model (Paper IV)
8Discussion and Conclusion
9Perspectives
10References
11Publications
Paper I:
“Animal models of DIC and their relevance to human DIC – A systematic review” (Submitted)
Paper II:
“Cardiovascular and haemostatic changes following microsphere injection in a rabbit model of acute pulmonary microvascular thromboembolism” (Submitted)
Paper III:
“Implementation of the ISTH classification of non-overt DIC in a thromboplastin induced rabbit model”. (Thrombosis Research, 2009; Vol. 124, Issue 4, Pages 490-497)
Paper IV:
“Purified thromboplastin causes haemostatic abnormalities but not overt DIC in an experimental rabbit model” (Thrombosis Research, 2010; in Press; DOI: 10.1016/j.thromres.2010.06.022)
Preface
“Whether you believe you can do a thing or not -you’re right”
Henry Ford
This thesis is the culmination of years of hard work, ups and downs and believing.
Early on in life I discovered my passion for the detail. I take every chance I get to learn and understand what is possible on a specific subject. However, the thought of doing a PhD did not settle in my mind until almost the finishing of my master studies (probably because I was too focused on the details of my master studies), but to be given the opportunity to take this detailed journey into PhD and DIC land has been fantastic. It has also been hard work - hard in other ways than I imagined. I expected it to be mostly scientifically challenging, but realized that working with my own response to set backs, frustrations and worries has been the hardest - and most rewarding.
“There are two kinds of problems; those you can solve - so don’t worry about them, and those you can’t solve - so don’t worry about them” Author unknown
I want to thank my colleagues at Novo Nordisk. Everywhere I went I was met by open minds and an eagerness to help. I specifically wish to thank my supervisors.Thanks to Henrik Duelund Pedersen (Novo Nordisk A/S), my supervisor for the first 1½ years, for teaching me focus on the experiments and the data and not worry so much about forms and conformities and for advancing me from student to colleague.Thanks to Annemarie (KU Life) for many high-level, high-speed meetings with extremely valuable professional inputs. Mikael (Novo Nordisk A/S) - in a rather chaotic switch from one department to another you quickly identified my strengths and weaknesses and used this constructively to create solutions where I could do my best. Thanks for always being there, being ready, listening and taking action.
My family and friends have had to put up with my absence in periods - especially during this spring, where the work has been most intense. Thank you for your patience.
And a special thank goes to my fiancé Martin and our son Viktor, who have felt the tough times as much as I have. Martin thank you for pushing, lifting and nursing me through the hard times of this PhD project - without you I wouldn’t have come this far (No, I wouldn’t !).
Abbreviations
ABPArterial blood pressure
APCActivated protein C
aPTTactivated partial thromboplastin time
ATAntithrombin
cTnICardiac troponin I
DICDisseminated intravascular coagulation
ECGElectrocardiogram
FDPFibrin degradation products
FVFactor V
FVa Activated Factor V
FVIIFactor VII
FVIIaActivated Factor VII
FVIIIFactor VIII
FVIIIa Activated Factor VIII
FIXFactor IX
FIXaActivated Factor IX
FXFactor X
FXaActivated Factor X
FXIFactor XI
FXIaActivated Factor XI
HEHaematoxylin eosin
ISTHInternational Society on Thrombosis and Haemostasis
LPSLipopolysaccharide
MIGETMultiple inert gas elimination technique
PaCO2Arterial CO2 tension
PaO2Arterial O2 tension
PAI-1Plasminogen activator inhibitor 1
PCProtein C
PSProtein S
PTProthrombin time
PTAHPhosphotungstic acid haematoxylin
RVPRight ventricular pressure
TATThrombin-antithrombin
TEGThromboelastography
TFTissue factor
TFPITissue factor pathway inhibitor
TMThrombomodulin
tPAtissue plasminogen activator
uPAurokinase type plasminogen activator
V/QVentilation-perfusion ratio
vWFvon Willebrand factor
Summary (English)
Disseminated intravascular coagulation (DIC) is a syndrome occurring secondary to a wide range of predisposing diseases. DIC is a pathological process with widespread thrombosis in the microcirculation causing organ damage and having a poor prognosis. Comparison of animal models of DIC to human DIC is crucial in order to translate findings in research models to treatment modalities for DIC in humans.
The aims of the current thesis were to establish a rabbit in vivomodel of thromboplastin induced DIC and evaluate the implementation of a wide panel of markers for pulmonary, cardiovascular and haemostatic function in the early identification of microthrombosis. It was hypothesized that a DIC scoring system can be applied to a rabbit model of thromboplastin induced DIC to standardize and score the rabbit model of DIC according to the ISTH scoring system of DIC in humans.
This thesis is composed of a general introduction to coagulation and its regulation and the syndrome DIC including an overview of animal models of DIC presented in paper I.
The core work of the project is then presented including the methods applied and results obtained.
These include the use of microspheres to establish a mechanical model of pulmonary embolism which is further described in paper II and a number of pilot studies developing a thromboplastin induced rabbit model of DIC with determination of a wide panel of markers for cardiovascular and haemostatic function including early markers of microthrombosis and finally characterising non-washed and purified thromboplastin. The validation and applicability of these methods are described in the set up of a non-clinical rabbit model of DIC in paper III and IV.
Finally the conclusion and perspectives are presented based on four accompanying papers.
Four papers are included in the thesis:
Paper I “Animal models of DIC and their relevance to human DIC – A systematic review”(submitted for publication) provides an extensive literature search on established animal models of DIC. An overview of distribution of animal species, inducers, measurements and treatments are given for the many identified studies. Generally a high variabilitywas found and it wasconcluded that standardization of animal models of DIC by for example application of a DIC score as has been developed for the diagnosis of human DIC by ISTH is recommendable. Furthermore it was identified that the majority of studies testing treatment of DIC only evaluated the prophylactic effect of such, though prophylaxis is in general irrelevant for cases of human DIC. The large amount of data compiled in this study is useful for the interpretation and comparison of responses in animal models of DIC and implementation of the recommendations may significantly improve the clinical relevance of animal models within this research area.
In paper II “Cardiovascular and haemostatic changes in a rabbit microsphere model of pulmonary thrombosis”(submitted for publication), microspheres were used to validate the response of cardiovascular parameters (RVP and systemic blood pressure) to fixed size pulmonary thrombi in the rabbit. Haemostatic parameters were evaluated in vivo and in vitro to test whether the microspheres are truly inert. A dose dependent response on cardiovascular parameters was observed in vivo, cumulating in a lethal effect of the highest in vivo dose of microspheres. An in vitro setup evaluating thromboelastographic effects of microspheres in rabbit whole blood spiked with equivalent doses to the in vivo study demonstrates that the highest dose resulted in a procoagulant effect. Haemostatic parameters showed no significant procoagulant effect of any of the doses of microspheres in vivo and it was concluded that non-lethal doses of microspheres result in inert pulmonary fixed size emboli.
Paper III “Implementation of the ISTH classification of non-overt DIC in a thromboplastin induced rabbit model”[1] describes the establishment of a rabbit model of thromboplastin induced non-overt DIC. Bolus injections of 1.25 and 2.5 mg thromboplastin/kg non-washed thromboplastin in rabbits resulted in non-overt DIC diagnosed by a modified score based on the ISTH score of non-overt DIC in humans, injection of higher thromboplastin doses were lethal and induction of overt DIC was not accomplished in this study.
In paper IV“Purified thromboplastin causes haemostatic abnormalities but not overt DIC in an experimental rabbit model” [2],it is reasoned that the purification of thromboplastin would decrease variability and enable injection of higher non-lethal doses resulting in overt DIC. However injection of a 2.5 mg/kg bolus followed by a 1.25 mg/kg infusion of thromboplastin resulted in less severe procoagulant activation than for the previous study, though a lethal effect was also observed within this study indicating an even more narrow window between procoagulant and lethal effect than for the injection of non-washed thromboplastin. These results underscores the difficulties in comparisons between animal models using different inducers of DIC and emphasizes that induction of DIC, experimentally as well as clinically, may call for different treatment approaches.
Sammendrag (Dansk)
Dissemineret intravaskulær koagulation (DIC) er et syndrom, der forekommer sekundært til en bred vifte af predisponerende sygdomme. DIC er en patologisk proces med udbredt trombose i microcirkulationen, hvilket forårsager organskade og giveren dårlig prognose. Sammenligning af dyremodeller for DIC med human DIC er altafgørende for at kunne translatere fund i forsøgsmodeller til behandlingsmodaliteter for human DIC.
Målsætningen for denne afhandling var at etablere en kanin in vivo model for thromboplastin induceret DIC og evaluere implementeringen af et bredt panel af markører for pulmonær, kardiovaskulær og hæmostatisk funktion i den tidlige identifikation af microtrombose. Hypotesen var, at et DIC scoringssystem kan anvendes i en kaninmodel for thromboplastin-induceret DICtil at standardisere og score kaninmodellen for DIC ifølge ISTH scoringssystemet for human DIC.
Denne afhandling omfatter en generel introduktion til koagulation og regulering af koagulationen og syndromet DIC inkluderende et overblik over dyremodeller for DIC præsenteret i artikel I.
Kernearbejderne i dette projekt præsenteres herefter og inkluderer de anvendte metoder og opnåede resultater. Disse inkluderer anvendelsen af microspherer til etablering af en mekanisk model for pulmonær embolisme, som beskrives nærmere i artikel II, samt et antal pilot studier med udvikling af en thromboplastin induceret kaninmodel for DIC med bestemmelse af et bredt panel af markører for kardiovaskulære og hæmostatiske funktioner inkluderende tidlige markører for microtrombose og endeligt karakterisering af ikke-vasket samt renset thromboplastin. Valideringen og anvendelsen af disse metoder er beskrevet i artikel III og IV.
Endelig præsenteres konklusionen og perspektiveringen baseret på 4 vedlagte artikler.
4 artikler indgår i afhandlingen:
Artikel I ”Animal models of DIC and their relevance to human DIC – A systematic review” (indsendt) omfatter et bredt litteraturstudie af etablerede DIC dyremodeller.Fordeling af dyrearter, induktionsmetoder, målinger og behandlinger oplyses for de mange identificerede studier. Generelt sås en høj variation i henhold til disse parametre og det blev konkluderet at standardisering af dyremodeller for DIC ved for eksempel anvendelse af en DIC score som tilsvarende er udviklet af ISTH til diagnosticering af human DIC anbefales. Ydermere bliver der vist, at hoveddelen af studierne, der tester behandling af DIC, udelukkende evaluerede den profylaktiske effekt af disse, selvom profylakse generelt er irrelevant for tilfælde af human DIC. Den store mængde data indsamlet til dette studie er brugbar for fortolkning og sammenligning af respons i DIC dyremodeller og implementering af disse anbefalinger kan signifikant forbedre den kliniske relevans af dyremodeller indenfor dette forskningsområde.
I artikel II ”Cardiovascular and haemostatic changes in a rabbit microsphere model of pulmonary thrombosis” (indsendt), anvendes microspherer til at validere det kardiovaskulære respons (RVP og systemisk blodtryk) til pulmonære tromber af fast størrelse i kaninen. Hæmostatiske parametre blev evalueretin vivo og in vitro for at teste om microsphererne med sikkerhed er inerte. Et dosis-afhængigt respons af kardiovaskulære parametre observeredesin vivo, kumulerende til en letal effekt af den højeste in vivo dosis af microspherer. Et in vitro setup, evaluerende tromboelastografisk effekt af den højeste in vivo dosis af microspherer i kaninfuldblod spiked med doser equivalent til in vivo studiet, demonstrerede at den højeste dosis resulterede i en prokoagulant effekt. Hæmostatiske parametre viste ingen signifikant prokoagulant effekt for nogen af microsphere-doserne in vivo og det konkluderedes, at non-letale doser af microspherer resulterer i inerte pulmonære emboli af fast størrelse.
Artikel III ”Implementation of the ISTH classification of non-overt DIC in a thromboplastin induced rabbit model” beskriver etableringen af en kaninmodel for tromboplastin-induceret non-overt DIC. Bolusinjektioner af 1.25 og 2.5 mg tromboplastin/kg ikke-vasket tromboplastin i kaniner resulterede i non-overt DIC diagnoseret af en modificeret score baseret på ISTH scoren for non-overt human DIC, injektioner af højere tromboplastin-doser var letale og induktion af overt DIC blev ikke opnået i dette studie.
I artikel IV ”Purified thromboplastin causes haemostatic abnormalities but not overt DIC in an experimental rabbit model” argumenteredes det, at rensning af tromboplastin ville mindske variationen og muliggøre injektion af højere non-letale doser resulterende i overt DIC. Men injektion af 2.5 mg/kg bolus fulgt a 1.25 mg/kg infusion af tromboplastin resulterede i svagere prokoagulant aktivering end i det tidligere studie, selvom en letal effekt også observeredes i dette studie, hvilket indikerer et endnu smallere vindue mellem prokoagulant og letal effekt end for injektionen af ikke-vasket tromboplastin. Disse resultater understreger problemerne ved sammenligning af dyremodeller, der bruger forskellige inducere af DIC og underbygger, at induktion af DIC, eksperimentelt som klinisk, kan kræve forskellige behandlingsstrategier.
1Introduction, hypotheses and objectives
Disseminated intravascular coagulation (DIC) is a complex and dynamic haemostaticsyndrome occurring secondary to a wide range of underlying diseases[3-5]. DIC is characterized by variable imbalances of the components of the coagulation and fibrinolytic systems and the clinical signs vary considerably ranging from no overt signs of disease (non-overt DIC)[6;7], accompanied by minor changes in haemostatic parameters, to clinical symptoms of organ failure, associated with micro-vascular thrombosis in vital organs, tofulminant DIC with bleeding symptoms (overt DIC) [6]. Diagnosis of DIC early in the non-overt stage may increase the chances of survival in the DIC patient, because early and aggressive intervention through supportive and antithrombotic therapy, besides treatment of the underlying disease, may minimize or prevent thrombo-embolic complications and progression to overt DIC[8].
Clinically relevant animal models of DIC are crucial in the research and development of novel therapeutic interventions as well as in understanding the pathogenesis of DIC. Standardization of animal models of DIC is necessary in order to compare findings between species and to increase the translational aspect in non-clinical testing of therapeutics for DIC in humans.
Simple diagnostic scoring systems for diagnosis of DIC in humans, such as the algorithms for overt and non-overt DIC developed by the International Society on Thrombosis and Haemostasis (ISTH) [7;9] are valuable standardized approaches to the characterization of DIC in humans and have been proven to have a high diagnostic accuracy [10-12]. As the approach taken by ISTH in scoring of DIC has been successfully applied in dogs suffering from DIC[13], application of similar scoring systems in experimental animal models of DIC may help to increase the clinical relevance and standardization of these non-clinical models.
The overall hypothesis of the ph.d.-study was that a DIC scoring system can be applied to a rabbit model of thromboplastin induced DIC. Since a similar scoring system for diagnosis of DIC in man has been validated as an accurate diagnostic tool, such an approach could lead to a standardized and clinical relevant animal model of non-overt and overt DIC.
Furthermore,it was hypothesized thatdetermination of several cardiovascular and haemostatic parameters would help to diagnose DIC in this model at an early stage and that the use of pulmonary function tests would add in the early diagnosis of microthromboembolism, an important complication of DIC.
The specific objectives of the proposed investigations were:
- To develop and standardize a non-clinical rabbit model of DIC with determination of a wide panel of markers forcardiovascular and haemostatic function including early markers of microthrombosis
- To investigate whether a rabbit model of DIC could be standardizedthrough application and scoring of the observed changes in laboratory markersaccording to the ISTH scoring system of DIC in humans
2Basic Mechanisms of Haemostasis and Thrombosis
Knowledge of basic mechanisms of haemostasisare important to understand the complexity of thrombo-hemorrhagic disease processes, to interpret laboratory testing modalities and to understand the mechanism of action of pharmacological agents being tested in non-clinicalanimal models.
2.1Physiologic Haemostasis
Coagulation is a complex process resulting in formation of clots. It is an important part of haemostasis, wherein a damaged blood vessel wall is covered by a platelet and fibrin-containing clot to stop bleeding and begin repair of the damaged vessel wall[14;15].
Haemostasis starts almost instantly after damage to the endothelium as exposure of the blood to proteins such as tissue factor (TF) initiates changes to platelets and fibrinogen[16]. Platelets immediately form a plug at the site of injury[17], simultaneouslycoagulation proteins in the blood plasma respond to form fibrin strands, which strengthen the platelet plug[18;19].