-. C¿- I S"' z..

Preyention of perioperative deepvein thrombosis in generalsurgery: a multicentre doubleblind study comparing two dosesof Logiparin and standard heparin

~

(

412-416

A total 011290 patients II'ere enrolled in a randomized multicel1tredouble blind study il1 arder to int~estigate the use of two doses 01 a 11ew101V molecular weight 11eparil1, Logiparil1@, in the pl~evention of deep veil1

¡ thron1bosis (D VI) il1 general. surgery. Patients who lI'ere illcluded Izadno col1traindication to 11eparill therapy al1d Izad at least Ol1e 01 therecogl1ized risk factors for D VT. Patients lI'ere randomized to receit~euírfi'actionatedheparil1 (UH) 5000 ul1its b.d., Logiparil12500 ul1its dailyor Logiparil13500 Ul1its daily. Each treat111ent lI'as git'el1 subcutal1eously2 h before surgery a.11d Col1til1ued for 7-10 days. Daily 125 /-Iabelled

fib,'inogel1 uptake tests (FUTs) IIJere peifor111ed fi.0111 day 2 to day 7 todet~ct D VT, and phleboangiography lI'as used to colljinn the diagl10sis.TI1e IVOUl1d 1vas e.\"aI11ined 011 a daily basis to check lor hae111ato111afon11ation, al1d GIl patiel1ts 11Jere folloll'ed up for 1 month alter operatiol1.All three treatmel1t am1s lI'ere well matched lor age, se.\", weigl1t,diagl10sis al1d type oloperation peiformed. The three majar il1Clusiol1criteria il1 the trial1vere 111aligl1ancy, age ot~er 6(J years and a 11istoryof varicose veil1s. Positive FUTs (UH=4.2 per cel1t, Logiparin2500 ul1its daily = 7,9 per cent, Logiparin 3500 uI1its daily = 3, 7 per cent)

al1d positive angiograms (UH =3,0 per cel1t, Logiparil1 2500 Ul1its~daily=5'6 per cel1t, Logiparin 3500 units daily=2.3 per cel1t) were¡Sigl1 ifical1 ti}' 1110re con1111011 il1 tl1e Logiparin 2500 units daily group than

il1 the UH al1d Logiparin 3500 Ul1its daily groups. The rates ofn1ajor

complicatiol1s (severe haelt10rrhage, death, pu{¡110nary en1bolism,reintervel1tiol1) 1..'ere silnilar in the three groups.

A, liezorovicz,H, Picolet,J, C. Peyrieux,J. P. Boissel and theH.B.P.M. ResearchGroup

Correspondence lo:Dr Alain Leizorovicz, HópitalNeuro-Cardiologique, Unité dePharmacologie Clinique,Départment Méthodologie etessais Therapeutiques, 162 avenueLacassagne, 69424 Lyon Cedex03, France

23 centres Jocated both in France and in the UK. Patients undergoinggeneral surgery (abdominal, gynaecological, urological or thoracic butnot cardiac surgery) who were 40 years or older and in whom generalanaesthesia longer Iban 30 min was anticipated were selected. Informedconsent was obtained in all cases. Only patients with at least one ofthe following Ti,k r,,{"jn,,-, rnr thTomboembolic disease were included inthe trial: previous historY~fvenous tnromboembOllSln, varicose veins,~(overweIght>20 per cent), C.S!ntracep,ljve Dill ~r hormQn~r~placementthe[aDY, chronic respiratorv insuQjciency, hSB.r1,-failllJ!.Jmalij(nancy, previous Ion!! bone fracture oflow~!Jirnp, bed rest > 5daY$b~for~ surg~ry, Ple2icted g_\l.ra!jQQQfJ~urge.!"y > 4~, age.~~ yea~.

The value oflow-dose heparin (5000 units b.i.d. or t.i.d.) for tlieprophylaxis of deep vein thrombosis (DVT) and pulmonaryembolism in general surgical patients has been extensivelyevaluated. A receni overview has confirmed the efficacy of sucha treatment which typically reduced the risk of DVT from22.4 per cent to 9 per cent and 9f pulmonary embolism from3 per cent to 1.7 per centl.

In the past few years new molecules of heparin with lowmolecular weight have been introduced. These have interestingnew biological properties and they have been aIleged in practiceto be superior to standard heparin (UH). Several compoundshave been developed and have been or are currently beingevaluated in patients2-8. An investigation was undertaken toassess the prophylactic value on DVT of a new low-dose lowmolecular weight heparin. Logiparin:K, Novo Nordisk,Copenhagen, Denmark as compared with a standard heparinregimen in middle/high risk patients undergoing major generalsurgery. In addition, since a major problem in clinical trialswith low molecular weight heparin is the choice of dose becauseno biological test aIlows a truly reliable indicator of itsantithrombotic action, two doses of Logiparin were comparedconcomitantly with UH. The protocol was accepted by theUniversité Claude Bernard Ethical Committee.

Sludy Irealmenls

There were Ihree Ireatment regimens in the tría!: Logiparin 2500units once a day (the preparations contained 8333 anti Xa units/ml),Logiparin 3500 units once a day (the preparations contained 11667anti Xa units/ml), and Heparin Novo 5000 units b.d. (sodium heparinprepared from hog intestinal mucosa). As Logiparin was given onlyonce dai]y a second injection was given daily (0.9 per cent NaCI) toensure that all three regimens were 'blind'. Two hours before surgicalintervention patients received a morning injection of the allocatedtreatment. The second injection was given 12 h later as an eveninginjection, and injections then followed every 12 h. AIl treatments wereadministered by subcutaneous injection. Treatment was contínued forat !east / days and for a maximum of 10 days. If thromboembolicprophylaxis was requíred for more than 10 days, UH was usedaccording to usual clínica] practice. Premature discontínuatíon of thestudy treatment was considered as a crítica! event and documentedaccording]y. Stockings or other forms of DVT prophylaxis were notallowed duríng the study periodo

Patients and methods

Paliem selecliollBetween June 1987 and November 1988, patients were recruited from

0007-1323/91/040412--OS if:) 1991 Buttcrworth-Hcínemann ltd412

Logiparinversus standard hepariningeneralsurgery: H.B.P.M. Research Group!

-

Main base/me characteristics .." ;,"'C, .", c,.;consldered posltlve, phlebography was performed wlthl~ 36 h of thetest.. ""';';';,""'"- -,:

¡'rabie:

~Logiparin: Logiparil1

Standard 2500units 3500 units

heparin.- daily daily(11 = 429} (11=431) (11=430)

AII angiogramswere reviewed cenlrally by Iwo inde~ndcnlradiologisls andthesiteoC DVT, ir present, was documcnlcd. Thcprotocols~cifiedthatvenous thrombosisconfirmed by angiographywasthe pnmary end.point:

10.0Biologica/. da.ta

Blood samples were drawn24 h before surgery (before the first studytreatmentinjection) and on day 3, day 5 and the day of discharge (butnot later than day 10). Red blood cell counts, plateJet counts,haematocrit, haemoglobin, activated partial thromboplastin time,prothrombin time andleucocyte counts were performed at each centre.Factor Xa inhibitory (XaI) activity was determined by an automatedchromogenic kinetic method at a centrallaboratory.

76.219.14.8

53.529.2

1.5

85.010.05.0

44.328.91.03.0

10.5

73:913-013.046.226.41.01.59-.1

Sample sizeThe total sample size wascomputed with the assumption of reductionin the incidence of venous thrombosis from 8 per cent (UH) to2.5 per cent (low molecular weight heparin group), as observed in aprevious study2, with type 1 error =0-05 and type II =0,05.

2.518.64.5

3-039-3

5,6

Risk (~~to~.Prevlous hlstory of venousthromboembolism

" .

Deepvenous thrombosls

PulmonaryembolismBoth "

"'. Varic.oseveiris-,,:,O,bCsrty (ov~rwelght>20%)c Contraceptive pills (current)

Hormonal therapy (current)Chronic respiratory

c insufficiencyHeart failureMalignancyPrevious long bone fracture

of.lower limbBed rest bCfore surgery

>5 daysExpected duration of surgery

>4hAgé ~60 yearsNumberof risk factors

9.4 16.9

Srorisricol onolysisAnalysis was first carried out on al! randomized patients, includingpatients who did not receive anystudy medication, in order to evaluateDVT as well as other end-points with the available information.Another analysis was performed onthe group of patients who receivcdat least one treatment injection. Univariate analyses were performedusing the classical statistical tests (Student's t test or analysis orvariance,as appropriate).

;6.4 ;8.4

29.932.822.911.0

29-233-221.812.4

3-00.52.3

31.538'1

18.38.13.11.0

4 ResultsPatients

Study population. A total of 1290 patients were randomlyallocated to receive either SH (429 patients), Logiparin2500 units daily (431 patients) or Logiparin 3500 units daily(430 patients}. Clinical and biological baseline characteristicswere fairly well balanced between the three groups; 45 baselinecharacteristics were compared and no biological or clinicalcomparison reached P = 0.05 level. There were 513 men and777 women and the mean age was 61 years (Table 1). Themea n number of risk factors for each patient at entry to thestudy was 2.3.

6Mean per patient

'pe of planned surgerAbdominal (n=921) 70,5

(303)14.4

(62)

72,6

(313)13,9

(60)9,3

(40)4.2

(18)

(305)174) 12,1

(52)11,2

(48)5,6

(24)

values in parentheses are num

lecologi<

,Jogic

(39)6-0

(26)

:rs of patients

Thoracic

68)

are percentagc Protocol deviations. Twenty-seven patients did not have anyFUT. Surgery was cancelled in seven and FUTs were notperformed in a further 20 (SH, n=8; Logiparin 2500 units daily,/1=8; Logiparin 3500 units daily, /1= 11). Al! 27 patients arenevertheless accounted for in the main analysis on an intentionto treat basis.

Disconlinualion of sludy lreal/11enl before day 8. In additionto the patients who did not receive any injection of studytreatment, 68 others prematurely stopped this treatment (oneof these patients was not operated on but received the firstpresurgical injection). The causes for discontinuation areshown in rabie 2.

Primary efficacy anal}'sislncidence of venous thrombosis confirmed by phlebography,

Positive FUT scans reported 10calIy le'd to phlebography in94 patients (four had two angiograms performed), Out of 64patients with a central positive FUT reading, three did nothave phlebography performed (beca use the patient or thesurgeon refused) and for six patients the films were lost by thelocal hospital and could not, therefore, be centralIy analysed.

The results of FUT scans and angiography as assessed bythe Central Angiography Reading Committee are shown inTables 3 and 4; no statistically significant difference betweenthe three groups was observed with respect to the incidenceof DVT at day 8. However, if aII types of venous thrombosisare accounted for (superficial andjor deep), there was asignificantly higher incidence of vein thrombosis in the

Study designA randomized double blind trial was organized by a CoordinatingCentre, a Steering Committee, a Critical Event Committee (whichassessed majar events) and an Executive Committee (operational armof the Steering Committee).

After baseline examination and assessment of eligibility, theinvestigator transmitted a request rOl allocation by computer networkto the Coordinating Centre. After checking eligibility the computerprogram allocated a treatment number corresponding to a treatmentbox containing the pre-allotted treatment.

Clinical examination was performed before inclusion, on day 8 afteroperation or at discharge, and 1 month later rOl evaluation ofend-points. Data collected (except documentatíon of critica! events)were circulated through a computer network. The advantages of sucha data system were absence of time lag due to mail and central keyprocesses, and on-line editing and correction.

Evalualion 01 end-poinls and clinical lollo~'-upA fibrinogen uptake test (FUT) was performed daily from day 2 to day7 or 8, On Sundays and bolidays tbe scan was performed only if apositive test was observedon tbe previous day; tbere should not bavebeen more tban I day missing in a series of scans for a given patientand tbe series sbould bave contained at least six tests if none was

positive,Two criteria were necessary for a positive scan: 20 per cent relative

increase between one point and the highest adjacent point or tbecontralateral point, and persistence of increase on the next day,Assessment of FUT was first done locally after wbich data from allpositive tests were reviewed centrally by a panel offour experts who wereunaware oftbe treatmenteacb patient bad received, Ira FUT was locally

413Br. J. Surg.. Vol. 78. No. 4. April1991

...'. cloglparln versus standard hepannln general surgery: H.B.P.M. Research Group

~.2Q

~~~91

1':+:-:::T

~~,í:;

x'1'0.05

~~

c" "'j"""",,,7:,'" .'Base,,' Day,3 ":;c:payj~,~", Dlscharge,'c "",","."c ,,'11", c

L6c'cJ.

group

BiologyAnti-Xa anal)'sis. AII the measurements were performedblindly in a centrallaboratory using an amidolytic techniquewith both SHand low molecular weight heparin standards. Atbaseline, factor XaI activity was similar in the three groups. Gnday 3, 3 h and 10 h after injection,on day 5 and on discharge,factor Xal activity was significantly higher in the low molecularweight heparin groups than in the UH group, and higher in theLogiparin 3500 units daily group than in the Logiparin

2500 units daily group (Figure 1).

, c!:'c': C ,cmore thal1 tO earsaoand th"¿/resentmanaemenrof sur ical

c- 'cy cgc,,"' c-Pc,c, g,c'y,.." g

patients, surgi,caf tech~iques, éarlymo~ilizatlo¡t:cand physio-:incide~ée,ofvenous

thromb<?s.ls s~en 111' the, present study.}n ~ore re:entrandomlzed ,tnals,companngUHand low molecular welghtheparins in general surgery, the incidence of positive FUT inthe SH groups rangéd from 11,9 percent to O per cent2-8.

Evaluation of the safety, of the treatment was one of thesecondary aims of the trial. Since haemorrhage is a majarconcern during anticoagulant therapy, one of the hoped forbenefits of low molecular weight heparin was a reduction inbleeding complications. So me studies have reported an increasein bleeding complications with low molecular weight heparin S,8.The doses might havebeen too highin these studies. The choiceof the appropriate clase should be made cautiously in view ofboth the results on DVT and possible sirle elTects. Whetherstandardization of the units of low molecular weight heparinwill help in clarifying the choice of the prophylactic clase is stilluncertain since no clear relationship between anti-Xa activityand efficacy as well as sirle elTects has truly been demonstrated.This trial is the first to address the question of the appropriateclase based on clinical end-points. However, practicallimitations due to sample size did not allow a study (lf a wider

range of Logiparin doses.In conclusion, the results óf this study demonstrate asignificant dilTerence in the efficacy of two clases of Logiparincompared with UH in the prevention of venouS thrombosis insurgical patients. Logiparin 2500 units daily was les s effectivethan SH and Logiparin 3500'units daily. The efficacies ofLogiparin 3500 units and UH were very similar and led to alow incidence of venouS thrombosis. The results are consistentwhether one looks at FUTpositive tests, any venous thrombosison phleboangiography or DVT on phleboangiography. Thethree regimens were well tolerated and there was no differencebetween the groups in the incidence of major as well as minarundesirable elTects. The efficacy and tolerance of Logiparin3500 units daily compared with UH are consistent with theresult observed in previous trials with other low molecular

weight heparins.

DiscussionThe present study was designed to evaluate the potential benefitof Logiparin in preventing DVT in patients after surgerycompared with low-dose standard heparin, to provideinformation about the best prophylactic dosage of Logiparinand to evaluate the safety ofthis medication. The results indicatethat there was no difference in the efficacy of Logiparin3500 units daily and SH 5000 b.d. in preventing postoperativeDVT while Logiparin 2500 units daily was significantly lesseffective than these two regimens. The initial assumption thatLogiparin could decrease the incidence of venous thrombosisas compared with SH could not be confirmed.

However, this trial differed i~ several aspects from othertrials assessing the prophylactic effect of heparin inperioperative patients. The diagnosis of DVT was made byFUT and confirmed by phleboangiography. Few trials haveadopted this definition for the main end-point. The reliabilityofboth methods has been evaluated by many investigators9-11.Although phleboangiography is still considered as the referencemethod, it may underestimate the occurrence of small transientthrombi, thrombi in formation, and, on the other hand, thrombiolder than 3 days. The advantages of phleboangiography arethe better reliability for femoral thrombi and the distinctionbetween deep and superficial thrombosis, while eliminating ralsepositive FUTs induced by haematoma or superficial thrombo-phlebitis. It algo has the greater advantage of permanentdocumentation, which can be reviewed whenever necessary.

To compare like with like, it seems fair to compare theincidence of positive FUTs observed in the present trial withthose reported in previous trials. In the SH group, the observedrate ofpositive FUT was 4.2 per cent which could be considereda low incidence. In a review of all trials comparing low-doseSH with controls by Collins et al.1, the overall incidence ofpositive FUTs in the SH group was 8,9 per cent (in 3966 generalsurgical patients). However, large variations from one trial toanother occurred. It should be stressed that in most of thesetrials a substantial number of randomized patients werewithdrawn from analysis, at least for the FUT end-point, forvarious reasons, and the effect might be to overestimate theincidence oí venouS thrombosis. Most trials were performed~

List of participants

Clinical centresDr Sagnard, Prof Dechavanne (biology), Dr White (isotopes),LyonService du Professeur Chabal, Hópital Edouard HerriotDr Combe- Tamzali, Dr Fretault (isotopes) ParísSef\'ice du Professeur Samama, Hópital Hotel DieuDr Parraguette, Prof Boneu (biology) ToulouseService du Professeur Lazorthes, C.H.U. PurpanDr Jarrige, ToulouseService du Professeur Gouzi, C.H,U. PurpanDr Rameil, NiceService du Professeur Bourgeon, Hópital Pasteur

Dr Perrier, ToursService du Professeur Brizon, C.H.U. Bretontteau

Dr Galland, AngersService du Professeur Soret, C.H.U.Dr Le Querrec, Dr Derlon (biology) CaenService du Professeur Gignoux, C.H.U.Dr Basse-CathaJinat, Dr Caix,PessacService du Professeur DucassoU, Hópital PelegrinDr Bastenaire, Dr Saint-Paul (biology) Le ChesnayCentre de Transfusion Sanguine, C.H.Dr Ballestrazzi, LilleService du Professeur Mazena. C.H.R.

Dr Dupont, LyonService du Professeur Tissot. Hópital Edouard HerriotProf Decousus, Dr Decousus,Saint EtienneService du Professeur Queneau, Hópital de BeJlevue

Dr Gille, LyonService du Professeur Partensky, Hópital Edouard Herriot

Dr Sayag, Dr Azzar (isotopes) LyonService du Professeur Braillon. Hópital Lyon Sud

Dr Mullet. LyonService du Professeur Dargent. Hópital Edouard Herri<>t

415

Br. J. Surg.. Vol. 78. No. 4. April1991