Minor events and the risk of deep venous thrombosis

(1)

Minor Events and the Risk of Deep Venous Thrombosis

E. M. W. Eekhoff, F. R. Rosendaal

1

, J. P. Vandenbroucke

From the Dept of Epidemiology and 'Haemostasis and Thrombosis Leiden University Medical Center Leiden, The Netherlands

Key words

Venous thromboembohsm. epidemiology, travel, tmmobihsation. nsk factors, factor V

Summary

Backgtound Deep venous thrombosis is a common disease, with genetic and acquired nsk factors Many patients have a history of minor events (short penods of iramobihsaüon such äs prolonged travel, short illness. ramor surgery or injunes) before onset of venous throm-bosis However. the role of these mmor events has received httle formal study Also, we do not know how mmor events might mteract with the presence of genetic prothrombotic defects (factor V Leiden mutation factor II mutation, protein C S and antithrombin deficiency) Patients and Methods On the basis of case-control data from a thrombosis Service m the Netherlands we added a follow-up penod for a case-cross-over analysis of minor events äs nsk factors, and a case only analysis toi the mteraction with factor V Leiden A total of 187 patients with first. objectively diagnosed venous thrombosis of the legs, aged 15-70, without underlymg malignanues and without major acquired nsk factors entered the study Tor the analy sis of minor events in the case-cross-over analysis we used a matched odds ratio, m the case only analysis we used the multiphcative s\nergy mdex Results In 32 6% ot the 187 patients with deep venous thrombosis who did not have major acquired nsk tactors, mmor events were the only external nsk factors Minor events mcreased the nsk of thrombosis about 3-fold äs estimated m the case-cross-over analysis (odds ratio 2 9, 95% con-fidence interval l 5-5 4) The synergy mdex between minor events and factor V Leiden mutation m the case-only analysis was 0 7 (95% con-fidence interval 03-15) Therefore, persons with factor V Leiden mutation who expenence a minor event will have an estimated nsk mcrease of about 17-fold. which exceeds the sum of the mdividual nsk factors Condusions Mmor events are hkelv to play an important role m the development of deep venous thrombosis, especially m the presence of genetic prothrombotic conditions

Introduction

The mcidence of deep venous thrombosis (DVT) m the general population is estimated to be l per 1000 per year Hereditary coagula-tion defects that are associated with an mcreased nsk are protein C, protein S, antithrombin deficiency, factor II (20210 G to A) mutation, and APC resistance (Factor V Leiden mutation) ( l ) Acquired condi-tions that promote DVT mclude prolonged immobihsation, major Correspondence to Prot Dr F R Rosendaal MD PhD Dept ot Climcal Epidemiology, Leiden Umversuv Medical Center, P 0 Box 9600 2300 RC Leiden, The Netherlands - Tel +031 71 5264037, Fax +031715248122. E mail rosendaal@mail medtac leidenumv nl

injunes and surgery The combination of acquired and hereditary nsk factors may lead to the highest nsks In a substantial number ot patients the underlymg cause remains unresolved, however

Many patients with DVT. when asked, have a recent history of mi-nor events stram or mmor injunes due to effort. short illness or short immobilisation while travellmg Smce the nsk brought about by major surgery and trauma, plaster casts and prolonged immobilisation is sub-stantial, mmor illnesses and mterventions might prove to be moderate nsk factors Nevertheless, mmor events have only received anecdotal attention m the hterature and very httle formal study except for (air) travel (2-4) Our study was performed to mvestigate the role of minor events m DVT, and their possible mteraction with genetic prothrom-botic defects We reanalysed data from the Leiden Thrombophilia Study (LETS) (5) in which mmor events, generally occumng within two weeks before the DVT, had been noted m patient discharge letters Firstly, we performed a case cross-over study four years after the first venous thrombosis patients were asked agam if thev had expenenced a minor event m a new and arbitrary two week penod This permitted us to use the cases äs their own controls Secondly we studied the svner-gy between mmor events and tactor V Leiden m a case only analysis

Methods

The present stud\ was an extension of the Leiden Thrombophilia Studv (LETS) (5) The LETS studv was pertormed between 1988 and 1993 and was a population based case control studv on hereditarv venous thrombosis Patients were selected trom the files of the anticoagulant clmics m Leiden Amsterdam and Rotterdam Patients with known malignant disorders were excluded In total 474 consecuti\e outpatients were mcluded (median age 47) Patients were referred tor diagnosis after a first objectivelv confirmed episode of DVT All patients were seen bv one of the mam mvestigators and completed a Standard questionnaire Se\eral major acquired thrombosis nsk tactors were recorded prolonged immobilisation surgery hospitahsation without surgerv pregnana and postpartum penod Genetic and biochemical nsk tactors analvsed initial ly or later on were tactor V Leiden mutation tactor II (20210 A/G) mutation and deficiencv ot protein C protein S or antithrombin

Although the LETS questionnaire did not melude Standard questions tor mmor events these were reguiarly mentioned äs a possible causattve tactor Most records trom the patients selected m Leiden contamed both extra comments ot the imestigator and hospital discharge letters The records trom Amsterdam and Rotterdam did not have this extra Information trom the discharge letters For our present study, theretore onlv the Leiden records were used

We selected patients without any major acquired nsk tactor, onlv patients without an e\ent or patients with a mmor event occumng within two weeks betöre the DVT were selected Mmor events were defmed äs bed rest more man 12 h j dav tor two or more consecutive days hcavy phvsical exer tion mmor trauma minor surgery or prolonged travel (by car bus or plane) leadmg to dimimshed or pamtul usc ot one ot the extrcmities

(2)

Eekhotf et αϊ.. Deep Venous ThromboMS

these patients, on average four years after the initial mclusion: patients were asked if a mmor event had occurred in the arbitrary two weeks penod before they had been approached for this follow-up The follow-up was by ques-tionnaire, augmented by telephone inquines: patients could mdicate on the questionnaire whether any mmor event had happened recently, and they were subsequently phoned to ask the details and time wmdow of these mmorevents Of each patient we therefore knew not only whether he or she had expenenced a mmor event m the (approximately) two weeks before the DVT, but also m an arbitrary two week penod at another time m their life This permits a matched case-control analysis to see how much more frequent such events are m the weeks preceding DVT than in another penod m the same patients life

Secondly. the mmor events preceding the ongmal venous thrombosis were studied m their combination with the presence of factor V Leiden The extern of mteraction between the exposure ot mmor events and the presence ot Factor V Leiden on the nsk for thrombosis was assessed by a case-only study (6, 7, 8) The case-only study relies on the assumption that the two factors ot mterest are mdependently distnbuted m the general population That is a reasonable assumption for genetic risk factors and mmor events for mmor events that precede a first DVT in a patient it is likely that their occurrence, i e whether or not the patient engaged m plane travel before a first DVT, was not mfluenced by the presence or absence of a hitherto unknown genetic factor (factor V Leiden and the prothroVnbm mutation was unknown at the time of our original investigation) DVT cases are divided mto those with and without mmor events The odds ratio of the tactor V Leiden is then an estimate of the synergy mdex on the multiphcative scale This synergy mdex is the factor by which the jomt relative nsks ot genetic defects and mmor events have to be multiphed Resuits

We mcluded 271 patients from the Leiden part of the LETS study The male/female ratio was 120/151, with mean age at diagnosis ot the thrombosis 42 years (ränge 15-69, SD 14) Of the 271 patients

84 were excluded because of pregnancy or postpartum. major opera-tions, prolonged immobilisaüon (more than two weeks), or hospital stay for more than three days preceding their DVT

Ot the remaimng 187 patients. 61 had expenenced mmor events (32 6%), the other 126 patients had no recorded event Table l shows the male/female ratio. mean age and ränge of the cases with <md without mmor events

Table 2 summanses the charactenstics of the mmor events The majonty was bed rest at home for vanous conditions, includmg mmor

Table I Sex and age charactenstics ot the 187 patients mcluded m the study male:female mean age + SD ränge

mmor events (n= 61) 3 1 3 0 43 3 ±13 4 1 4 6 - 6 8 7 no mmor events (n= 126) 5 5 7 1 42 8 +.14 6 1 5 3 - 6 7 3

Table 2 Charactenstics of the mmor events preceding DVT

Minor event Patients Remark

Table 3 Coagulation abnormaliües at time of deep venous thrombosis

Travel 1 1 Rest at home> 1 2 h 36

Mmor surgery 14

car 6, plane 3, bus l, otherwise l hernia 6, malaise 6, mfection 2, trauma 10, heavy physical exer-tion 12

arthroscopy 6, memscectomy 2, excision ganghon/pin knee 2, ex-cision exostosis l, sclerosmg of vancesl, coronary angiography l, other l Coagulation abnormality protein C protein S FV Leiden AG FV Leiden AA Factor II (202 10 A)

with minor event (n=61) 1 2 9 . -no event (n=126) 7 2 22 3 6

Table 4 Resuits of the case cross-over studv

Mmor Events present before DVT absent before DVT

present m recent penod absent m recent penod

9 35 44 12 72 84 21 107 128

mjunes Travel and mmor surgery had lesser contnbutions The median mterval between the events and the thrombosis was 2 days (ränge 0-14) The median time of the duration of the mmor event was l day (ränge

l-14 days)

Coagulation abnormaliües are summansed m Table 3 In 12 of the 61 patients (20%) with a mmor event, a hereditory abnormality of coag-ulation was found (9 patients had a factor V Leiden) In the patients without minor events, 40 of the 126 patients (30%) had a Coagulation abnormality (25 factor V Leiden) Note that 3 patients who were homo-zygotes for factor V Leiden and most of the patients with a protein C deficiency were found m the patient group without minor e\ents

The case-cross-over study that was pertormed at approximately four years after the initial event. dunng a follow-up investigation. is present-ed m Table 4 Of the 187 patients, 128 could be mcludpresent-ed Several reasons existed for the exclusion of the 59 patients, ranging from pregnancy, nontraceabihty major medical e\ents or unwillmgness to participate Of these 128 patients. 21 had expenenced a minor event dunng the recent two week penod preceding the time ot follow-up The matched analysis of the presence or absence of mmor events is present-ed in Table 4 The matchpresent-ed odds-ratio äs an estimate of the relative nsk. is 2 9 (confidence mterval l 55 4)

We can restnct the case-cross-over analysis m Table 4 to those without factor V Leiden, m order to measure the mdependent effect of mmor events Twenty of the 128 patients camed factor V Leiden all were among those without a recent mmor e\ent (6 ot 35 and 14 of 72) The matched odds ratio without these patients became 2 4, which is

Table 5 Charactenstics of the mmor events m case cross-over study

Recent mmor event Patients Type of recent event Mmor event at original

(3)

Thromb Haemost 2000 83 408-11

Table 6 Results of the case-onlv study

Minor events FV Leiden + FV Leiden

-+ 9 25 34 52 101 153 61 126 187

sufficiently dose to the overall estimate of 29, given the small numbers

The type of minor e\ ents durmg the more recent two week penod is presented m Table 5, together with the original minor event of the same groups of patients Conspicuously absent from the hst of recent minor e\ents m companson with the ones precedmg DVT is minor surgery The relative nsk for minor surgery would therefore be very high

The synergy mdex of the mteraction between minor events and factor V Leiden was obtamed by a case-only analysis on the original data of DVT i e , the minor events precedmg DVT and factor V Leiden DNA determmations In an analysis on Table 6, the odds-raüo. estimat-mg the synergy mdex, is 0 69 with a confidence mterval of 0 31 5, which mdicates the presence of an mcomplete multiplicative effect of about 0 7

Discussion

Our results support the hypothesis that minor events play an im-portant role m the cause of \ enous thrombosis To estimate the separate effect of minoi events, we used a case-cross-over study m which the cases w ere their own controls Four years after the venous thrombosis, patients were asked agam if they had expenenced minor events in an arbiträr}' two-week penod This yielded a close to 3-fold mcrease m nsk of venous thrombosis after minor events In addition we calculated a synergy mdex of about 0 7 between small events and factor V Leiden If we take the mdependent nsk elevation of factor V Leiden to be 8-fold this means that the overall relative nsk of a minor event m a person who carnes the factor V Leiden mutation will be 8 times 3 times 0 7, which amounts to a relative nsk of about 17 That jomt relative nsk is clearly higher than the effect of each nsk factor alone, and exceeds the sum of the relative nsks

Before accepting these estimates at face value, we have to acknowl-edge potential strengths and weaknesses m our design The ongmal LETS study was not designed for the aim of this investigation, the association with minor events, however, was spontaneously noticed by the physician or mvestigator One third of the patients had a minor event m the weeks precedmg their DVT, which is a remarkable amount This confirms the overall Impression of physicians that patients often volunteer a history of minor events To study how high the frequency really is, we used the patients äs their own controls, and asked them about similar minor events dunng an arbitrary 2 week penod at the occasion of a four year follow-up investigation Smce it is possible that the ongmal spontaneous recording by physicians was elicited by some m-depth probmg of the patients, we also probed rather m-depth dunng the follow-up patients could mdicate on a questionnaire whether any of the small events had been happemng recently, and they were sub-sequently phoned to ask the details and time wmdow of these small events. Of course, our study is possibly hampered by the increased age and consequent changes in hfe habits of the patients. We did not obtain follow up on all patients, although we do thmk it unhkely that participation in the follow-up would be associated with the expenence

of a minor event Nevertheless, a bias could also be mtroduced m the patient group due to their famihanty with the possible nsk of minor events

A number of studies with a few cases tends to support the hypothesis that minor events play a significant role m DVT We found at least twenty reports m the hterature that studied the relation of thrombo-embohsm and minor events, pubhshed between 1940 and 1998 (9-29) Most papers were descnptions of cases or case senes and concerned pulmonary embohsm, some with detectable DVT, mamly associated with air travel Dunng air travel cabin related nsk factors leadmg to hypercoagulability and hypoxia are postulated to mcrease the nsk to develop DVT due to stasis Only few papers also descnbed the asso-ciation with other traffic (car, tram, ship), immobilisation in theatre/ cinema, or due to effort, stram or cramped conditions in shelters (among others 9 21, 24 27) There were few relevant larger studies One retrospective case senes identified 44 patients with DVT related to air flight (from a larger retrospective senes of 254 persons), of which a minonty without known nsk factors, but also some with mjury to the lower extremity pnor to travel (20) A recent formal case-control study with 160 cases and the same number of controls, mvestigatmg all modes of travel and gomg back several weeks m the history of the patient, found that 24 5% of patients and 7 5% of controls had a history of travel of more than five hours, leadmg to a fourfold mcrease in nsk (4)

In our study the jomt effect of factor V Leiden and minor e\ ents was shown to be high, at a 17 fold mcrease in nsk of the combmation This means that additional cases of DVT occur due to the synerg) of minor events and factor V Leiden, which would otherwise not have occurred by either nsk factor alone One should also bear m mmd that several nsk factors might enhance each other, e g factor V Leiden, minor events and oral contraceptive use In a recent study m patients takmg oral contraceptives who developed deep-vem thrombosis or pulmonar) embohsm, the factor V Leiden and acquired nsk factors (both minor and major events) were both found to be multiplicative nsk factors over and beyond oral contraceptive use (30) Unfortunately, the relative contnbution of minor and major events could not be estimated from that study These results are important, smce a recent epidemic of pulmonary embohsm deaths with third generation contraceptives in New Zealand concerned a number of women who had expenenced minor events dunng use of these contraceptives» (31) Rather than blaming the minor e\ ents, it becomes mcreasmgly clear from our study and several others that the concomitant use of more thrombogemc contraceptives leads to the heightened nsk (32,33)

In conclusion, our study confirms the clmical notion that minor events are hkely to play a role m the etiology of DVT, especially in the presence of hereditor} thrombogemc nsk factors

Acknouledgement

We acknowledge the great help of the patients who \olunteered for the initial study and the follow-up exammation We also acknowledge the efforts of Dr T Koster and of our data-managers Ms W Noteboom and Ms A Schreijer m organismg the initial studv and the follow-up The ongmal study was supported by a grant from the Netherlands Heart Foundation

References

l Lane DA, Mannucci PM, Bauer KA, Bertina RM, Bochkov NP, Boulyjen-kov V, Chandy M, Dahlback B, Gmter EK, Miletich JP, Rosendaal FR, Seligsohn U Inhented thrombophilia Part l Thromb Haemost 1996, 76 651-62

(4)

bekholl et al Doep Venous l hrombosis

2 Houghton A, Taylor P Vascularhazardsofairtravel BntishJClmPractice 1993,47 60-1

3 Milne R Venous thromboembolism and travei is there an association J R C o l l Phystcians Lond 1992, l 47-9

4 Ferrari E. Chevalher Th, Chapelier A, Baudouy M Travei äs a nsk factor for venous thromboembolic disease Chest 1999, 115 440-4

5 Koster T, Rosendaal FR, de Ronde H, Bnet E, Vandenbroucke JP, Bertina RM Venous thrombosis due to poor anticoagulant response to activated protemC Leiden Thrombophiha Study Lancet 1993, 342 15036 6 BeggCB ZhangZF Statistical analysisof molecularepidemiology studies

employmg case senes Cancer Epidemiology Biomarkers & Pre\ention 1994,3 173-5

7 Khoury MJ, Flanders WD Nontraditional epidemiologic approaches m the analysis of gene environment mteraction case control studies with no controls1 AmJEpidemiol 1996, 144 207-13

8 Piegorsch WW, Weinberg CR Taylor JA Non-hierarchical logistic models and case only designs for assessmg susceptibihty m population-based case control studies Statistics m Medicme 1994 13 153-62

9 BenoitR Trav eller thromboembolic disease the economv class syndrome J Mal Vase 1992 17 B84-7

10 Beighton PH, Richards PR Cardiovascular disease m air travellers Bnt HeartJ1968 30 367V

11 Sarvesvaran R Sudden natural deaths associated with commercial air travei Med Sei Law 1986 26 35-8

12 Paganin F Laurent Υ Gauzere BA Blanc P Roblm X Pulmonary embo

hsm on non-stop flights between France and reunion island Lancet 1996 347 11956

13 Lederman JA, Keshavarzian A Acute pulmonary embolism followmg air travei Postgr Med J 1983 59 104-5

14 Marshall M Flugreisen-Thrombose Munch Med Wochenschr 1982, 124 83

15 Finch P Ransford R, Hill Smith A Thromboembolism and air travei Lancet 1988 u 1025

16 Hollidav J Atvpitdl presentation ot multiple puimonarv emboli m a young traveller J R Coll Gen Pract 1985 35 497

17 Homans I Thrombosis of the deep leg vems due to prolonged bittmg New Engl J Med 1964 250 148-9

18 Black J Deep-vem thrombosis and puimonarv embolism Lancet 1993 342 3^2 3

19 Cruickshank JM, Gorlm R, Jennett B Air travei and thrombotic episodes the economy class syndrome Lancet 1988, π 497-8

20 Kistner RL, Eklof 80, Masuda EM, Sonntag BV, Wong HP Venous thromboembolism m association with prolonged air travei Dermatology Surgery 1996, 22 637-41

21 Simpson K Shelter deaths from pulmonary embolism Lancet 1940 2 744 22 Steinhauser RP, Stewart JC Deep venous thrombosis in the mihtary pilot

Aviat Spac Environ 1989 60 1096-8

23 Symmgton IS Stack BHR Pulmonary thromboembolism after travei Br J Dis Chest 1977, 71 138-40

24 Crane C Deep venous thrombosis in leg followmg effort strain New Engl J Med 1952, 246 529-33

25 Cairns NJW, Thomas JEP Abson CP Pulmonary embolism a hazard of air travei Cent A f r J Med 1981 27 85-87 1981

26 Burki U Lungenembohen bei und nach Lang-Streckenflugen (economy class syndrome) Schweiz Med Wochenschrift 1989, 119 287-9 27 Medier RG McQueen DA Effort thrombosis m a young wrestler J Bone

Jomt Surgery 1993 75 1071 3

28 O'Donnell D Thromboembolism and air travei Lancet 1988 n 797 29 Mercer A, Brown JD Venous thromboembolism associated with air travei

a report of 33 patients Aviation Space and Emironmental Medicme 1998 69 154-7

30 Schambeck C Schwender S Haubitz I Geisen UE Grossmann RE Keller F Selective screenmg tor the Factor V Leiden mutation is it ad\isable pnor to the prescnption ot oral contraceptives7 Thromb Haemost 1997 78

1480-3

31 Medicmes Adverse Reaction Committee of New Zealand URL www medsate govt nz/hprots htm', go to Medicmes Safety', Third generation contraceptives and pulmonary embolism December 1998 and February 1999

32 WHO Scientific Group Cardiovascular Disease and Steroid Hormone Contraception WHO Technical Report Senes 877 WHO Geneva 1998 33 Walker AM Newer oral contraceptives and the nsk of venous thrombo

embolism Contraception 1998 57 169-81

Received June 23 1999 Accepted atter rewbion October 28 1999