Interleukin 8 and venous thrombosis: evidence for a role of inflammation in thrombosis

Interleukin 8 and venous thrombosis: evidence for a role of

inflammation in thrombosis

Rosendaal, F.R.

Citation

Rosendaal, F. R. (2002). Interleukin 8 and venous thrombosis: evidence for a role of

inflammation in thrombosis, 173-177. Retrieved from https://hdl.handle.net/1887/1591

Version:

Not Applicable (or Unknown)

License:

Downloaded from:

https://hdl.handle.net/1887/1591

Bniitih Journal of Haematologii 2002 116 173 177

Interleukin 8 and venous thrombosis: evidence

for a role of inflammation in thrombosis

BIMI N E \ \ \ ^ PH n R H Ri π s\i \ ' \ \ u FRI ι s R Rosi SD \ \ i 5 4 lLaboiatonj toi I \peunu ntal Intciual Mediane and 2Depaitment of Vasculai Mediane Academu Medual Centeι Amsteidam and 'Depaitnunt oiHanatoloqij

and 4Oepartment of Chmcal Epidemwlo0i] Leiden Umveisiti) Medual Centn Leiden Thi \cthuhiuds

Received 19 July 2001 accepted foi publualion 37 Auqmt 200Ί

Summary. Elevdted plasmd levels of interleukm 8 (II-S) were previously shown to be associated with recurrcnt venous thrombosis To dssess the nsk öl venous thrombosis

IL-8 plasmd concentrdtions were measured in pdticnts dnd control subjects of the Leiden Thrombophüid Stud\ ( I I T S ) This populdtion bdsed cdse-control stud\ mcluded 474 pdtients with d iirst deep-vem thrombosis cind 474 age- and sex-matched controls The nsk oi venous thiombosis for subjects with elevdted IL-8 levels (dbovc 90th peicentile of controls) compared with subjects with IL-8 levels below the 90th percentile was mcrcased l 8-fold (95%C1 12-28)

Adjusted ioi age and sex tht odds ratio w d s 19 (95' iCI l 3-2 S) II 8 concentrations w e r e weakh coireldted with dge male sex and concentrations of C-reactive protein icictor \ III codgulation activiU and homocv steine but adjustment lor these lactors did not substdntialh affect tht dssocicition between II-8 dnd venous thrombosis Our results suggest that II-8 is a nsk fdctoi for venous thiom bosis

Keywords venous thrombosis epidemiolog\ interleukm S inflammation nsk factors

Classic risk factors for venous thrombosis mclude acquired factors such äs immobih/ation surgerv and mahgnancies and genetic risk factors i e activated protein C resistance (factor V Leiden) and dehuenues of protein C (PC) protein S (PS) or antithrombm (AT) Recently elevated plasma levels of factor VIII hyperhomocystemaemia and prolhrom-bm 2021OA have been added to the serics oi risk fcictois (reviewed by Rosendaal 1999) A previous studv in paticnts with recurrent venous thrombosis suggested inlcrleukm 8 (IL-8) to be d nsk fdctor for venous thrombosis (Vdn Aken (i al 2000)

Interleukm 8 d C-X-C chemokme is produced b> several cell types mcluding endothehal cells pcnpheral blood monocytes neutrophils epithehal cells dnd hbioblasts (reviewed by Hoch dal 1996 dnd l u s t e r 1998) The mdin function of IL-8 is the dctivation oi integnn-mediated adhesion of neutrophils (luster 1998) and it was recenth shown thdt IL-8 is a powerful tngger foi adhesion of monocytes to vascular endothehum (Gers/ten et al 1999) Several studies have identifled IL-8 in association w i t h various acute and chronic inflammatorv tonditions includ

Correspondence 0r P H Reitsma l abordtorv foi l xpenmentril Internal Mediane Atddemit Medicdl C c n t u G2-129 M c i b c i i d i c i i 9 1105 A/ Amsterdam I h c Nelhtrlands l mail P H Rutsmdfo drnt uvti nl

mg sepsis psonasis rheumatoid arthiitis and asthm i (reviewed bv H o t h d n J 1996) P i o i n l l a m m a t o i v i M o k n u s IL I ß dnd tumoui neciosis fattoi ~J. ( I M y) and v n a l and bacteridl Stimuli um m d u i t II S p i o d u i t i o n

T h e impoitante öl i n f l a m m a l o n i t s p o n s t s in \ t n o u s thrombosis hds Ixen shown in m \iin> studies (Johnson ( i «i 1996 Senden dal 1 9 9 S I and in ammal modtls (\Aakelield < ( al 1 9 9 5 1995 1 9 9 7 ) bul l i t t l c is k n o w n aboul the associalion of i n f l d i n m a t o n mediators and \ c n o u s thiombosis m humans Studies usmg a h u m a n t \perimcntal cndotoxatmia modtl h a v e suggested cm

interaclion between intlammatorv mediatois and blood toagulation ( V a n üeventer ( i al 1990) Infusion öl endo toxm m hcalth\ \olunteeis resulttd in mueased plasma levels oi I I - l β II-6 II 8 dnd TM α äs well ds of coaguldtion actixation markers Recenth elevated ton centrations oi C l e a c t u e prolein ( C R P ) h a v e been detetted in patients w i t h a hrst cpisode öl v e n o u s thrombosis suggestmg an assotiation with svstemic mllammation (Kamphuisen et al 1999)

Based on thesc consideration5 v\e theoii^ed that II 8

levels mav pla\ a c a u s a t n e role in v t n o u s thiombosis In oidei to obtain evidenct for this supposition the associdtion bttween elevdted plasma levels of II 8 and venous throm-bosis w a s analvsed in a tast-tontrol studv öl pdtients with a In st episodt öl vcnous thrombosis

174

PATIEN FS A N D METHODS

populatwn I he I eiden Thrombophihd Study (LLTS)

hds been descnbed previouslv (Koster et al 1993) Bneily 474 patients with d hrst episode of objectivcly demonstraled deep venous thrombosis dnd 474 dge and sex-matched hedlthy controls were included Pdtients with known nidligndnt disorders were excluded dnd pdüents were seen

only dlter dnticodguldnt trcatment had been discontmued for dt ledst 3 months Ihis Weis dlwdvs > 6 monlhs alter the event because oral anticoagulant (OAC) tre.ilment Weis routmel·} given tor 3 months I orty-eighl ( 1 0 % ) o! ihe pdtients were on long-term coumann trealment and were not dllowed tu Interrupt their medicdtion tor vanous reasons (Koster et al 1995) The healthy control subjects were acquamtances oi pdtients or partners öl other patients and were selected according to the lollowing cnteria sdme sex same age (± 5 \ e a r s ) no biological relationship no historv öl venous thromboembolism no use oi coumdrin-denvdlives lor dt least 3 months and no known malignanues

lalmmtoni sfm/ii's Blood was collected Irom the

ante-cubital vein mto tubes contammg 0 106 mol/I tnsodium utrate Plasnid wds prepdred by centnlugdtion lor 10 min dt 2000 g dt room temperature dnd stored at — 70°C 1L-8 concentrations were measured usmg an en/vme linked immunoborbent assav (LL1SA) with d detection hmit öl 2 0 pg II-8/ml (Central Laboratory of the Nietherlands Red Cross Blood Transfusion Service CLB Amsterdam l he Netherlands) Plasma samples were assaved in d blind manner If the dillerente between the mean öl the duphcate measurements and the measurements sepdrdtely exceeded 10% of the medii vdlue the medsurement wds not repeated and the sdmple wds excluded from furlher dndlvsis Fhis resulted in d case/control ratio of 466/462 Samples below the detection hmit were designated äs 2 pg/ml

SinfisüKiii (innii/sn We calculated odds ratlos äs an estimdte öl the relative nsk öl venous thrombosis m subjects with elevated IL-8 concentrations usmg the 90th percentile öl IL-8 äs a cut-otl value based on the distnbution in the control population and under the assumption that levels measured alter the event tdirly represent those of before the event The odds rdtios in the group with elevdted IL-8 concentrations were calculdted usmg logistic regression dnalvsis with the group below the 90th percentile äs the reference category All odds rdtio cdlculdtions were adjusted for dge dnd sex except when stratidcation lor sex or age was perlormed The 90th percentile ot the IL-8 concentrations of the control study population was 7 l pg/ml for men it was 7 5 pg/ml and for women it was 6 8 pg/ml fhe possible conlounding effect öl age on the association between IL-8 and venous thrombosis was analysed usmg equally distnbuted age-tertiles designated young (lowest tertile) middle (second tertile) and old (upper tertile) Tor the total study population the age tertiles were < 40 40-51 and > 51 vears

RESl LTS

Ihe mean age of the whole population was 45 years (ränge 15-72) 57% were women Ihirty-six per cent öl

both patients and control subjects were current smokers while the mean body mass index (BMI) was 26 (ränge 14-46) kg/mr The mean II -8 concentration wds higher m the patients than the controls [8 0 pg/ml (95%CI 60-100) versus 4 4 pg/ml (95%CI 4 0 4 8 ) ] Lleven per cent of the controls and 9% of the pdtients had pldsma levels below the detection hmit (2 0 pg/ml) The IL-8 plasma concen-trations öl individual patients and controls are shown in

rig i

Ιο evdludte the association between IL-8 plasma con-centrations and venous thrombosis a cut-off value öl

7 l pg/ml (90th percentile of the control population) was used Seventy-seven ( 1 7 % ) patients were detected with elevated IL-8 concentrations (above 90th percentile) com-pared with 45 (10% by defimtion) of the controls fodds ratio l 9 (95%CI l 2-2 8)J Usmg the cut-off of 162 pg/ml (99th percentile) 2 3 ( 5 % ) patients with elevated IL-8 were delected in companson with iour (1% by defimtion) controls [odds ratio 60 (95%CI 20-17)] Odds ratios usmg other cut-oll values are shown in Table I By dividmg the IL-8 concentrations mto qumtiles and companng each qumtile with the lowest we lound a concentration-dependent dssocidtion pomting towards d threshold effect The nsk Wds not mcreased in the second [odds ratio 10 (95%CI 07-1 6)] dnd third qumtile [odds ratio 10 (95%CI 06-1 5)] while in the lourth qumtile [odds ratio l 2 (95%CI 0 8-1 8)] and m the fifth qumtile [odds ratio l 8 (95%CI 1 2-2 8)] the odds ratio was mcreased

In control subjecls we exammed which factors mfluenced the 11-8 concentration The mean IL-8 concentration was higher in men [5 3 pg/ml (95%CI 4 4-6 1)] than m women [3 8 pg/ml (95%CI 3 4-4 1)J IL-8 concentrations mcreased by 0 2 pg/ml (95%CI-0 5-1 0) per 10 years of age Distri-bution of the control subjects in age groups showed that 30 (10%) öl the subjects between 40 and 51 years and 70 ( 2 3 % ) above the age öl 51 years had an elevated IL-8 concentration (> 7 l pg/ml 90th percentile) compared with 22 ( 7 % ) of the subjects under the age of 40 years

fhe mean IL-8 concentration was higher in subjects wilh elevated (> 5 l μg/ml 90th percentile) CRP plasma

levels [5 5 pg/ml (95%CI 30-80)] than m other subjects [4 3 pg/ml (95%CI 39-47)]

Risk lactors for venous thrombosis include high levels öl

lactor VIII coagulant activity (FVIII C) and homocysteme Among healthy subjects with elevated FVIII C concentra-tions (> 100 lU/dl) the IL-8 concentration was slightly mcreased [n = 344 4 6 pg/ml (95% CI 4 0-5 1)] compared with subjects with low ΓΥΙΙ1 C (< 100 IU/dl)[n = 130

Intel leukm 8 and Venous Thiombosis

175

IL-8 (pg/ml)

100

10

i ig l Plasma levels of II -8 dre elevated m p itients with venous thrombosis compared \ \ i t h control subjects The solid lines lepresent t ho mean the dotted hnes the detection hmit I ho numbers at the top of the cases column indicdte the IL-8 levels that were abovc

100 pg/ml in hve patients

l

1

'*"»!»»«>""

CASES

Table I Thrombosis nsk for different cut oft pomts of IL S concentrdtions

IL-8 plasma concentrdtions Pdtients (466) Controls (462)

Odds ratio adjusted for dge dnd scx

O R I 9 5 U ) < 80th peicentile ( 5 6 pg/ml) > 80th percentile < 90th percentile ( 7 1 pg/ml) > 90th peioentilo < 95th peicentile (8 2 pg/ml) > 95th percentile < 99th peroentile ( 1 6 2 pa/ml) > 99th percentile 336 130 389 77 401 65 443 23 3 7 3 S 9 417 45 440 22 45S 4 1 7 ( 1 2 - 2 1 9 ( 1 2 - 2 3 3 (20 i 6 0 (2 0 1 3 ) S) 5 l 7)

non smokers [respectivelv 4 2 mg/1 (95%CI 5 8-4 7) versus 4 5 (95%CI 39-5 1)]

Next we investigated the lole of the dbove-mentioned factots on the assouation between elevated IL-8 (> 90th percentile) and venous thrombosis Ad]ustment for FVIII C [odds, ratio l 7 (95%CI l 1-2 6)] homocvsteme [odds ratio l 7 (95%CI l 1-2 6)] CRP [odds ratio l 8 (95%CI l 2-2 6)] or BMI [odds ratio l 7 (95%CI l 1-2 6)1 only margmalh affected the assocwtion between venous thrombosis and elevated IL-8 plasma concenlrations

The presence of other thrombosis nsk faotors (factor V Leiden prothrombm G20210A iaotor VIII faotor IX tactor XI and antithrombm protein C 01 protem S defioiencies) did not dflect the association between elevated II 8 (> 9()th percentile) and venous thrombosis [odds ratio l 7 ( 9 5 ' X C I l 1-2 7)] and neither did exolusion öl these patients jodds ratio l 9 (95'XCI l 2-2 9)]

Patients wert mcludcd m the stud\ at different time intervals after the thrombotit event [mean 21 (ränge 6-68) months] II 8 conoentiation w a s not reldted to this time mtervcil Patients who were seen more than 28 months alter the thrombotio event had a simildr frequenc\ (16'X ) of ekvated IL S concentration äs those seen mithin 12 months after the thrombosis ( 1 8 Ή ) Addilionallv anahsis after

exolusion of the sub]ects vuth an elevated CRP concentra tion (> 5 l μg/ml 90th poicentile) did not affeot the etssotidtion between IL S and \cnous thrombosis although 51 ( Π Χ ) patients vuth elevaled IL-8 tonoentidtions uere present comparcd with 37 ( 9 / 1 of the controls [odds l a t i o

l 6 (95'XCI l 0-2 5)|

II-S oonoentrations were highor among oial dnttooagu lant (OAC) useis than among non usois l he reasons lor continuins; OAC tieatment \\cio äs lollous 15 of theso 4S

patients had A l e o u i r e m o si\ hdd a histon of m\ooardial

176 Β Γ wtii Aken P H Rcit\ma and l R Rowndaal

iniarction (XII) and seven nere diagnosed vuth PC/PS/A l dciiuencv I hus 28 out öl 48 mdividuals had a clear

medical reason tor continumg OAC treatment Ihese condilions (recurrence MI second thrombophilic delect) \\ere prcsent in onl\ 19 öl ihe other 399 patients

rwelve out öl 46 ( 2 6 / > ) patients still usmg OAC had elevaled II S concenlrations in contrast to 65 ( I S / M öl thc 420 patients vuthout C)AC Lxclusion of OAC users resulted in 6 (16%) patients vuth elevated 1L 8 concentrdtions (> 90th percentile) versus 45 (10 K.) controls [l 7 (95%CI l 1-2 5)1 l sing the cut-otf öl 7 l pg/ml (90th percentile) the adjusted odds i d t i o was 2 9 (95%CI 2 1 8 - 4 9 ) and usmg the cut-off of 16 2 pg/ml (99th percentile) the adjusled odds ratio was 6 0 (95%CI 2 0-1 7 7)

Separate analysis showed simüar odds ratios for men and women vuth elevated [L 8 concentrations |for men l 5 (95 (Cl 0 8 - 2 6 ) and Ior women 2 2 (95%CI l 5 - 5 9 ) ] Stralihcation bv age showed that the effect of II-8 was hmited to middle-aged subjects [age 50-51 vears odds ratio 5 4 (95%CI l 6-72)1 while m younger [age < 50 odds latio 29 (95".CI 06-147)1 or elderly subjects [age > 5 1 \ears odds lalio l 5 (95%CI 0 8 - 2 2 ) ) an mcreased nsk öl venous thrombosis was less obvious

D I b C L S S I O N

Elevated concentrations öl II-8 aie assouated with venous thrombosis Plasma concentrations oi II-8 (above the 90th percentile) lead to a l 9-lold (ad]usted ior age and sex) mcreased nsk of venous thrombosis The assoctalion between venous thrombosis and IL-8 is most pronounced between the age öl 40 and 51 vears and is not atfected by levels ot ΓVIII C or homocvsteme

Scveral mechamsms may explain ihe association between elevated II-8 concenlrations and venous thrombosis First IL-8 has been shown to mduce tissue lactor an important mducer ot blood coagulation on monocvles (\eumann

et al 1997) Second leucocvte recruitment is induced by

IL-8 (Luster 1998) Leucocytes are the ftrsl cells to adhere to venous endothehum in a rnodel ot stasis-induced deep venous thrombosis and these leucocytes then sttmulate thrombus formation (Schaubci al 1984) Furthermore IL-8 could contnbute to the mduction of a procoagulant surface by triggenng the adhesion oi monocvtes to the endothehum (Gers/ten et al 1999)

Ihe main stimuh for IL-8 production are proinflamma-tory cytokmes such äs IL-1 and Τ\Γ-α bactenal products and viral miecnon (Baggiohm ei «I 1994) It can only be speculated which Stimuli are responsible Ior Ihe mcreased IL-8 concentrations associated with venous thrombosis The potential role of mfectious agents m the pathogenesis of vascular disease has been studied m relation to atherosclerosis These studies have suggested that mfec-lions with tor example Chlamijdm pneumomae Helnobacter

pylori and cytomegalovirus are associated with

atheroscle-rosis (Ross 1999) These mfectious agents may mduce the production oi IL-8 m the endothehum and monocytes (Kragsbjerg et «l 1995 Bliss et al 1998 Murayama et «l 1998) Therefore it is worthwhile to explore the potential

role öl Ihese miecüons m the mduction öl IL-8 m relation

to venous thrombosis

Studies in baboons have shown elevations in IL-8 levels alter mduction oi a thrombus suggestmg that the thrombotic event could contnbute to ein increased IL 8 concentralion (Wakefleld et al 1995) Nevertheless it is unlikely that the thrombosis itselt was responsible for the mcreased IL-8 concentrations äs the subjects were mcluded at least 6 months alter the thrombotic event IL 8 concentration was not related to the ttme mterval between thrombotic event and measurement of IL-8 and no decrease in IL-8 level was detected in patients who were mcluded more than 28 months after thetr throm-botic event compared with the patients mcluded wtthm 12 months Although plasma levels ot CRP a marker Ior inflammation are high in patients ad]ustmenl for CRP did not ai'fect the thrombosis nsk ot elevated acute phase protems FVIII C and flbrmogen (Kamphuisen et al 1999) or IL-8 suggestmg that these assoctations are not owing to a post-thrombotic acute phase response However we cannot exclude the possibility that post-thrombotic syn-drome or underlymg malignancies contnbuted to the elevated IL-8 concentration in some patients which may in pari or completely be responsible for the observed dssociation

Another possibility is that the mcreased concentrations ot IL-8 reflect occult cancer because many cancers produce IL-8 It is also well-known that patients with idiopathic venous thrombosis have a higher frequency of underlymg cancer (Monreal et al 1997) In fact venous thrombosis is often the first sign ot mahgnancy m most patients with prostatic and pancreatic carunoma The design of the LETS study excludes patients with known cancer and overt cancer does not therefore explam the results The LETS protocoi however did not mclude an exhaustive search for occult cancer and non-svmptomatic cancer may therefore explam all or part of the mcreased IL-8 concentrations Only large prospective studies focusmg on cancer durmg iollow-up will provide a defimte answer

In conclusion the association between elevated IL-8 concentrations and a first thrombotic event provides evidence for the notion that IL-8 mcreases the nsk oi venous thrombosis

ACKNOWLEDGMENTS

We thank Dr T Koster ior collecting blood samples öl patients and controls Dr F J M van der Meer (head Thrombosis Centre Leiden) Dr L P Colly (head Thrombosis Centre Amsterdam) and Dr P H Tnenekens (head Thrombosis Centre Rotterdam) ior their kind cooperation and Mrs T Vissei and Ms A P Groot for sküful technical assistance

REFERENCES

Bdggiolmi M Dewald B & Moser B (1994) Interleukm S aml related chemotactic cytokines - CXC and CC chemoluiu-s

Advaiuts in Imnwnologit 55 97-179

Intel leukm 8 and Venous Thwmbosis

177

Hl,ss |r C M Golenbock D T Ktates S Lmevskj } K & K i l h t P (1998) Hehcobacter pjlon hpopolysacchande bmds ι 01)14 and stimulates release of mlerleukin 8 epithehal luuliophil dctivdtmg peptide 78 and monocyte chemotacüc

piolun l by human monocytes Injeitwn Immumty 66 3357-i > h 5

ι n s / t t n R E Garcia/epeda t A lim Y C Yoshida M l)mi, H A Gimbrone Jr M A Luster A D Luscinskas F W & Roscn/weig A (1999) MCP l and IL 8 tngger firm ddhesion of monocytes to vasculai endothelium under Ilow conditions \ K i i i i 398 718-723

i l u h R C Schraufstattei I U & Cochrane C G (1996) In vivo in utio and molecular dspects of mterleukm 8 and the inter luikin 8 receptors Journal of Laboratonj Chnual Mediane 128

l 54-145

Johnson K Aarden L Choi Υ De Groot E & Creasey A (1996) Iht promflammatory cytokine response to coagulation and indotoxm m whole blood Bioorf 87 3031-5060

k imphuisen P W Elkenboom ] C ] Vos H L Pablo R Sturk A Btrtma R M & Rosendaal FR (1999) Increased levels of factoi \ III and fibrmogen m patients with venous thrombosis are not c msed by acute phase reactions Thrombosis and Hnemostnsis 81 680-683

koster 1 Rosendaal F R de Ronde H Briet E Vandenbroucke J P & Bertina R M (1993) Venous thrombosis due to poor anticoagulant response to activated protun C Leiden Ihrombo philidStudy Lernet 342 1503-1506

koster T Rosendaal F R Briet E van der Meer Γ J Collv L P Tnenekens P H Poort S R Reitsma P H & Vandenbroucke J P (1995) Protein C deficiency m a controlled senes of unselected outpahents an mfrequent but clear nsk factor for venous thrombosis (Leiden Thrombophiha Study) Blood 85 2756-2761

kragsbjerg P Jones I Vikerfors T & Holmberg H (199b) Diagnostic value of blood cytokine concentrations in acute pneumoma Thora\ 50 1253-1257

luster A D (1998) Chemokmes - Chemotactic cvtokines that mediale mflammation New England Journal of Medium 338 436-445

Vlonreal M Fernandez Llamazares ] Perandreu ] l rratia A Sahuqmllo J C & Contel E (1997) Occult cancer in patients vuth venous thromboembolism which patients which cancers

Thrombosis and Haemostasis 78 1316-1318

Murajama l Mukaida N Khabar K S 6, Matsushima k (1998) Potential imohement of IL 8 m the pathogenesis of human cv tomegalovirus infection Journal öl l eitkot Ute Bwloqi/ 64 62 67

Neumann F J Ott l Marx N Luther l Kenngott S Gawa/ M Kot?sch M \Schomig A (1997) Fffcct of human recombmant mteileukm 6 and mterleukm S on monocyte procoagulant activity /Irtenosiieiosit T/iiomhosis and \asiiilai Bioloqii 17

3399-34Ü5

Rosendaal I· R (1999) Venous thrombosis a multicausal disease

lamel 353 H6"-1173

ROSS R (1999) Atherosclerosis an mflammaton disease V»

England Journal o/ Medium 340 1 1 5 126

Schaub R C Simmons C A Koets M H Romano P J 6, Stewart G ] (1984) Faih events m the foimation of a \enous thrombus followmg local tiauma and stasis Inboiatonf inv<sti

qntwn·! 51 21S-224

Senden N H M Jeunhomme I \1 A A Htemskeik J U M \ a n Wagenvoord R Van t Veer C Hemker H ( _ 6. Buurman \\ \ (1998) Factoi Xa induces c\tokme production and expiession oi adhesion molecules b\ human umbihcal \em endothehal cells

Joiunal öl Immunoloqii 161 4318-4324

Van Aken B l· Den Heijer M Bös G M J Van De\ enter S ] H 6, Reitsma P H (2000) Rccurrcnt venous thrombosis and maikeis of mflammation T/irom/wsis and Hiidiiostiiiii 83 336 539 Van Devcntu S J Bullti H R ten Cite J V\ Λ irden I \

Hack C Γ i. Sturk A (1990) Ixpeiimental endotoxcmia in humans anahsis of cvtokmt releast and coagulation libimo l>tic and complemcnt pathwavs Bioorf 76 2:>20-2->26 Wakefield T V\ Greenfaeld L I Rolfe M V\ Dtlucia \ Sti itter

R M Abrams G D Kunkel S I Ismon L1 \\ robltski S K Kadell AM Burdick MD & Tenno l (1993) Inflammaton and procoagulant mediator interactions m an experimental baboon model of venous thrombosis T/HOHI/IOSÜ and HiKmostosis

69 164-172

Wakefield l W Stnctcr R M V\ ilke C A Kadell A M Wrobleski S K Burdick M D Schmidt R Kunkel S l & Greenheld L J ( 1 9 9 3 ) Venous thrombosis associated mflamma tion and cittcnuation w i t h ncutrahzmg antibodies to cvtokines and ddhesion molecules Irdnosdoosis T/iimubosis and \iiMulai

Bwloqy 15 238 268

Wakelield T W Stncter R M Puncc M R Dowmng I J 6. Greenheld L] ( 1 9 9 ~ ) Pathoaenesis of venous thrombosis a nev\ msight Cwdw\asiulai Sntyeni 1997 6 1 5