Plante, J., Boneu, B., Vaysse, c., Barret, A., Gouzi , M. & Bierme, R. (1979):

Dipyridarnole-aspirin versus low doses of heparin in the prophylaxis of deep venous

thrombosis in abdominal surgery. Thromb. Res., 14,399-403.

Renaud, S. & Godu, J. (1970). Thrombosis prevention by acetylsalicylic acid in hyperlipemic

rats. Can. med. Ass. J., 103,1037-1040.

Renney, J. T. G., O'Sullivan, E. F. & Burke, P. F. (1976). Prevention ofpostoperative deep vein

thrombosis with dipyridamole and aspirin. Brit. med. J., 1,992-994.

Rogers, P. H., Walsh, P. N., Marder, V. J., Bosak, G. c., Lachrnan, J. W., Ritchie, W. G. M.,

Oppenheimer, L. & Sherry, S. (1978). Controlled trial of low-dose heparin and

sulphinpyrazone to prevent venous thromboembolism after operation on the hip. J. Bone

Joint Surg., 60, 758-762.

Rosenberg, F. J., Gimber-Phillips, P. E., Groblewski, G. E., Davison , c., Phillips, D. K.,

Goralnick, S. J. & Cahill , E. D. (1971). Acetylsalicylic acid: Inhibition of platelet

aggregation in the rabbit. 1. Pharm. exp. Ther., 179,410-418.

Roth, G. J. & Majerus , P. W. (1975). The mechanism ofthe effect ofaspirin on human platelets.

J. clin. Invest ., 56,624-632.

Salzman, E. W., Harris, W. H. & DeSanctis, R. W. (1971). Reduction in venous

thromboembolism by agents affecting platclet function . New Eng. J. Med., 284,

1287-1292.

Schöndorf, T. H. & Hey, D. (1978). Modified " low-dose" heparin prophylaxis to reduce

thrombosis after hip joint operations. Thromb. Res., 12,163-163.

Sevitt , S. (1973). Pathology and pathogenesis of deep vein thrombi. In Pulmonary

Thromboembolism. cd. Moser, K. M. & Stein, M., p. 93 . Chicago: Year Book Medical

Publishers.

Silvergleid, A. J., Bernstein , R., Burton . D. S., Tanner, J. B., Silverrnan , J. F. & Schrier, S. L.

(1977). Aspirin-Persantin prophylaxis in elective total hip replacement. Thromb.

Haemost., 38, 166.

Smith, J. B. & Willis, A. L. (1971). Aspirin selectively inhibits prostagiandin production in

human platelets. Nature New Bioi.. 231,235-237.

Smythe, H. A., Ogryzlo , M. A., Murphy, E. A. & Mustard, J. F. (1965). The effect of

sulphinpyrazone (Anturan) on platelet economy and blood coagulation in man . Can . med.

Ass. J., 92, 818-821.

The Canadian Cooperative Study Group (1978). A randomized trial of aspirin and

sulphinpyrazone in threatened stroke . New Eng. J. Med., 299, 53-59.

Weily, H. S. & Genton, E. (1970). Altered platelet function in patients with prosthetic mitral

valves. Effects ofsulphinpyrazone therapy. Circulation, 42,967-972.

Wylie, J. S., Chesterman. C. N., Morgan , F. J. & Castaldi, P. A. (1979). The effect of

sulphinpyrazone on the aggregation and release reactions of human platelets. Thromb.

Res ., 14,23-33.

Wu , T. K., Tsapogas, M. J. & Jordan, F. R. (1977). Prophylaxis of deep venous thrombosis by

hydroxychloroquine sulfate and heparin. Surg. Gynec. Obst., 145, 714-718.

VENO US T HROMBOEM BOLlSM AND ANTIPLATELET DRUGS 233

Zekert, F., Kohn, P., Vormi ttag, E., Poigen furst, J. & Th ien, M. (1974). Thromb oembolic

Proph ylaxe mit Acetylsalic ylsaure bei Operat ionen wegen huftgelenknaher Frakturen.

Monatschrift der Unfa llheilku nde. 77, 97-110.

Zucker , M. B. & Peterson, J. (1970). EfTect of acetylsalicylic acid, other non-stero idal

anti-inflarnmatory agents, and dipyridamole on hu man blood platelets. 1. lab. c1in. Med..

76, 66-75.

BASIC CURRENTEVIDENCE FOR

THEUSE OF DRUGS TO BOTH PREVENT

AND TREAT PATIENTS WITH

VENOUSTHROMBOSIS AND

PULMONARY EMBOLISM

c. V. RUCKLEY

The Royal Inftrmary, Edinburgh EH3 9YW. Scotland

This paper will largely be devoted to the role of platelet suppressant agents in the

prophylaxis of venous thrombosis in surgieal patients. It is worth pointing out ,

however, that the universal interest in prevention and in aeute thrombosis somet imes

obseures the faet that venous disease offers a number of other areas of therapeutie

challenge. These include the prevention of recurring thrombotie and /or embolie

episodes , the avoidanee or the amelioration of the post-thrornbotic syndrome - a

largely neglected disorder - and the need to prevent thrombosis after venous

reconstruction, perhaps the most difficult area of all. Whether agents which interfere

with the interaction between blood cells and vessel wall have apart to play in any or

all ofthese situations is as yet far from clear.

Venous thromboembolism remains a eommon cause of death after operation.

Surgery provides a unique opportunity to study thrombosis because we surgeons

ereate, as an ineidental accompaniment of operation, the ideal circumstances which

Virchow prescribed for thrombogenesis - alterations ofblood flow, in the direction of

stasis or turbulence, vessel wall lesions and haematological disturbance. Moreover,

since the onset ofthe thrombosis can be antieipated, in the rnajority, to within a day

or two, the process can be monitored from its inception to its outcome with noninvasive means such as the ['l5I]-fJbrinogen test. We know the probable natural

history ofvenous thrombosis in surgical populations if not in individuals.

It is not surprising therefore that the prevention of post-operative venous

thrombosis has been such a prol ific research field in the last decade . However, agents

whieh intervene direetly in the interaction between the vessel wall and the circulating

platelet or white eell are the most recent candidates for evaluation, aspirin apart; they

have been studied more in arterial than venous disease, and information is sparse as

compared with agents which act predominantly on the coagulation system .

In considering the response to prophylactie agents in different clinieal situations, it

is important to bear in mind that more than one meehanism may be operating in the

early stages ofthrombus formation .

Experimental thrombosis has usually been induced by injury to the vessel wall;

direct trauma, laser, ehemieal, thermal or e1ectrieal injury . Here the first event is

DRUGS IN VENOUS THROMBOSIS AND PULMONARY EMBOLlSM 235

adherence of platelets to exposed collagen and their local aggregation to each other.

The platelet release reaction leads to stabilisation of white thrombus and binding

with fibrin (Welch, 1887; French, 1965). Platelet suppressant therapy might therefore

be effective prophylaxis to cover operations in which there is likely to be direct

vascular damage such as orthopaedic, venous or arterial procedures, particularly in

the pelvis or lower limbs.

The mechanism ofnidus formation where there is no direct vascular trauma is less

cIear. Most thrombi begin in the veins ofthe lower leg, in calfrnuscles, sinusoids or in

venous valve sinuses with relative anoxia (Malone, Hamer & Silver, 1979)and stasis

as probable trigger factors. Stasis alone, certainly in the short terrn, does not give rise

to venous thrombosis, but is a potent factor in combination with a hypercoagulable

state which may pre-exist or be induced by the stress and trauma of surgery. Wu &

Man sfield (1980) have recently shown in pigs that stasis by pneumatic cuff leads to

detectable changes in endothelial cells within three minutes and desquamation

within ten minutes. They also observed the damage to be most marked in and around

the valves .

The experimental simulation ofthe stasis thrombus is the WessIer model in which

injected thrombin-free serum, to produce a transient hypercoagulable state, is

combined with stasis (Wessler, 1955). The thrombus thus produced is not a laminated

platelet and fibrin structure such as is found in tlowing blood but a mass of red cells

interlaced by a fme network containing scattered platelets and leucocytes in the same

proportions as in blood. When such a thrombus comes into contact with tlowing

blood fibrin, leucocytes and platelets selectively accumulate and adhere to its surface.

Olson, Lungquist & Bergentz (1974) analysed radiolabelIed platelets and fibrin in

experimental arterial and venous thrombi formed in dogs by the insertion of steel

tubes into vessels and showed that platelet content was 68 times higher in arterial and

28 times higher in venous blood than in normal blood. The fibrin content was

approximately twice the fibrinogen concentration of plasma. When however the

animals were made hypofibrinogenaemic by reptilase almost no thrombi were

formed in either arteries or veins . They concIuded that fibrin was very important in

early thrombus formation (Olson, 1974).

Early thrombi in humans have been studied in autopsy specimens. Paterson (1969)

reported on 26 valve cusp thrombi mostly from femoral veins - the majority were

white thrombi containing platelet masses. Sevitt (1974)provided conl1icting evidence

in a study of 48 such valve pocket thrombi. He found that at their apices -

presumably the starting points - such thrombi were formed predominantly of fibrinred cell collections. The difference may be important.

Markers ofthe platelet release reaction, such as ß-thromboglobulin , can be shown

to be elevated in early thrombosis as diagnosed by the (125I]-flbrinogen test (Smith,

Duncanson, Ruckley, Webber, Allan, Dawes, Bolton, Hunter, Pepper & Cash, 1978),

but the latter, although the most sensitive of diagnostictests, cannot be cIaimed to

identify initial nidus formation.

Evidently then in different experimental situations the initiating role ofthe platelet

is not always apparent, the structure ofthrombus is different and the proportions of

the cellular constituents vary . It seems reasonable to suppose that the same may be

true ofdifferent cIinical situations.

Although attention has focused mainly on platelets, the leucocyte may have a

crucial role in thrombogenesis. Migration of white cells across the vessel wall has

lang been recognised as part of the response of tissues to sublethai injuries. Stewart,

Ritchie & Lynch showed in 1974 that the combination of stasis and distal trauma

resulted in leucocyte mediated endothelial damage. Cooke, LIoyd, Bowcock &

Pileher (1977)subsequently showed in a controlled trial that intravenous lignocaine,

which inhibits leucocyte adhesion and invasion, significantly reduced the incidence

of phlebographically proven deep vein thrombosis after elective hip surgery. This is

236 c. V. RUCKLEY

an a rea of prophylaxis which does not seem to have been taken up by other workers.

Pro p hylaxis can be divided into mechanica l and pharmacologica l methods. Even

simple m echanical methods such as electrical calf muscle stimulation or the more

popular intermittent pneumatic compression, both of which have been shown by

controlled trials to reduce the incidence of (12SI]-flbrinogen-proven deep vein

thrombosis, must influence the interaction between the endothelial cell and blood

constituents by altering the velocity ofblood flow , an illustration ofthe difficulty of

attempting completely to categorise prophylactic agents as acting on one particular

component ofVirchow's triad.

T his difficulty is further illustrated when we consider the various pharmacological

approaches. T hese can be subdivided into the use of drugs which act as anticoagulants, those which stimulate fibrinolysis, and those which suppress platelet

activity, The two major prophylactic agents in current use , low dose hepari n and

dextran, fit mainly into the anticoagulant group, a lthough both mayaiso inhibit

platelet ad hesion and aggregation. T hese two have been extensively studied in trials

of prophylaxis, are of proven va lue (except for the failure of heparin in hip surgery),

and wi ll not be discussed further.

Agents which stimulate the endothel ia l ce ll to re lease plasmi nogen ac tivator, suc h

as stanozalol or ethy loes trenol and ph en formin , altho ugh theoretically attractive,

have proved disappointing. Fossard , Friend, Field, Corrigan, Kakkar & Flute (1974)

found in a double-b lind randomised st udy that treatment with et hy loestrenol and

phenformin three weeks before and one week after elcctive gynaecological surgery

had no effect on the inciden cc of post-operative venous thrombosis detected with th e

['2SI]-fIbrinogen test. If age nts can be develope d which will enhance fib rinolytic

activity over a long period of'time, without inducing bleeding, and are weil tolerated

in other respects, this could still be a fruitfu l approach.

Platelct suppressant agents which have been tested in clinical trials in surgical

patients include aspirin, dipyridamolc, su lphinpyrazone, flurbiprofen and hydroxyTable 1 The effectiveness of some platelet suppressant agents against deep vein thrombosis

(DVT) and pulmonary embolism (PE) following general surgery.

N umber

0/

Author patients Diagn osis Regimens DVTI PE2 Signiftcan ce

M.R.C. 303 1251Ff3 Aspirin 600 mg daily for (27.5%) N.S.

Steering 5 days (153 patients)

Committee Control (150 patients) (22.0%)

(1972)

Clagett 104 12SIFf Aspirin 650 mg twice daily 7(12.5%) N.S.

et al. (56 patients)

(1974) Control (59 patients) 10(20.4%) - Renney 160 1251FT Aspirin 1000mg + 12(14%) P<0.02

et al. dipyridamole 100mg

(1976) daily (85 patients)

Control (75 patients) 24(32%)

Loew 187 I2SIFf Aspirin 500 mg i.v. twice 19(30%)

et al. daily (63 patients)

(1977) Heparin 5000 units s.c. 11(19%) 2

twice daily (57 patients)

Aspirin + Heparin 5(9%) P < 0.008

(57 patients)

Numbe r of patients developing DVT.

Number of patients developing PE.

J [ 1251]-flbrinogen test (12SIFf).

N.S. Not significant,

DR UG S IN VENOUS T HROM BOSIS AND PULM O NARY EMBOLlSM 237

chloroquine . In general surger y, when we are dealing predominantl y with stasis

thrombi, th e res ults as su m marised in Table I have been unrewarding as far as

aspirin, in a dose of600 mg dail y or twice daily, is concerned (Renney, O 'Sullivan &

Burk e, 1976 ; Loew, Brocke, Simma, Vinazzer, Di en st yl & Boehmne, 1977 ; MRC

Ste ering Committee, 1972 ; C lagett, Brier, Rosoff, Schneider & Salzman, 1974) .

Renney et al. (1976) in a contro lled tria l in pat ients undergoing abdominal surgery

demonstrated a benefit wh en asp irin and dipyridamole were combi ned. This require s

confirrnation. In an Austrian study (Lo ew et al., 1977) aspirin plus heparin were

significantly more effec tive th an eithe r agent a lone , and surpris ingly no bleeding

problem was reported. In orthopaedic surgery the situation may be different.

Table 2 Incidence of deep vein thrombosis in patients following orthopaed ic surgery.

(Collected series).

Fractured femur

Elective hip/knee

Mean incidence

50%

54%

Range

35-75 %

47-58%

Reference

Salzman & Harris

(19 76)

The incide nc e of deep vein th rombosis (DVT) (Table 2) in orthopaedic patients is

high and pulmonary embolism th e commonest ca use of death (Salzman & Harr is,

1976 ). A wide variet y of regimens have been tried in fractured femur (Table 3)

(Wood, Pr entice, McGrouther , Sinclair & McNicol , 1973 ; Morris & Mitch eII, 1977 ;

Table3 The effectiveness of various drug regimens against deep vein thrombosis (DVT) and

pulmonary embolism (PE) in patients with fractured femur.

Nu mber

0/

Author patients Diagnosis Regim ens D VT ' PE 2 Significance

Wood 30 '251FT] RA 233 +aspirin 600 mg 9(100%)

et al. daily (9 patient s)

(1973) RA 233 +placebo 11(92%) N.S.

(12 patients)

Placebo (9 patients) 6(67%)

Morris& 176 1251FT Control (24 patients) 16(67%)

Mitchell Heparin 5000 units three 12(50%) N.S.

(1977) times daily (24 patients)

Dipyridamole 100mg three 15(63%)

times daily (24 patients)

Control (32 patients) 21(66%)

Aspirin 300 mg 20(63%) N.S.

three times daily

+dipyridamole 100mg

three times daily

(32 patients)

Control (20 patients) 12(60%)

Flurbiprofen (20 patients) 13(65%) N.S.

Rogers 30 1251FT Control (16patients) 12(75%)

eta!. Heparin 5000 units twice 5(36%) P<0 .04

(1978) daily+sulphinpyrazone

200 mg four times daily

(14patients)

Numb er of patient s developing DVT.

Number of patients developing PE.

) 1251fibrinogen test.

N.S. Not significant.

238 c. V. RUCKLEY

Rog ers, Walsh, Marder, Bazak , Lachman , Ritch ie, Oppenheim er & Sherry, 1978).

The universal lack ofsuccess, with the possible exception of heparin combined with

sulphinpyrazo ne (Ro gers et al., 1978) ma y be attributable not so much to the

ine fficacy of the drugs as to th e fact th at th e vein damage is incur red and th e

thrombotic process generall y establ ished before treatment is begun . It is unlikely that

any regime n will substantially reduce the risk of th romboembolism in patients with

fractured fem ur without at th e same tim e incurring a significant risk ofbleeding.

Hip reconstruction op er ations hav e been th e commo nest field for tr ials of plat elet

suppression (Tables 4 and 5). Salzm an , Harris & de Sanctis (1971) compared

dipyridamole, aspirin, warfarin and dextran. Di agnosis was by clini cal means. The

dipyridamole gro up was stopped at 34 becau se 9 pati ents developed thromboembolism ; the other three showed sim ilar levels of DVT. Th e study was unsati sfactory in that th e cont ro l group was retrospecti ve.

Table4 The effectiveness of various platelet suppressant agents against deep vein thrombosis

(DVT) and pulmonary embolism (PE) in patients following hip reconstruction I.

Nu mber

0/

Author patients Diagnosis Regim en PE' DVP S ignifica nce

Salzman 169 Clinical Dipyridamole 3 6(18%) > 0.1

et al. 100mg four times

(1 971) daily (34 patients)

Aspirin 600 mg 4 4(9%) < 0.02

twice daily (43 patients)

Warfarin (43 patients) 2 4(9%) < 0.02

Dextran 40, 500 ml 2 5(1 0%) < 0.01

daily for4 days,

500 ml every 3 days

(49 patients)

Retrospective controls 7 23(34%)

(67 patients)

2 Harris 187 Phlebography Aspirin 600 mg twice 0 18(35%) < 0.01 eta!. daily (51 patients)

(19 74) Warfarin (55 patients) 0 10(1 8%) < 0.01

Dextran 40, 500 ml 2 14(23%) < 0.01

daily for 3 days,

500 ml every 3 days

(61 patients)

Heparin 5000 units 11(55%)

three times daily

(20 patients)

I Number of'patients developing PE.

2 Number ofpatients developing DVT.

Harris, Salzman, Athanasoulis, Waltman, Baum & de San cti s, in 1974, using

phlebography to di agnose, again had no untreated control group but showed aspirin,

warfarin and dextran to be signi ficantly better th an low do se heparin. The reduction

ac hieved by asp irin was not impressive.

Dechav anne, Ville , Viala, Kh er, Faivre , Pousset & Dejour (1975) in a rat her small

study found a higher rate of DVT in th eir group treated with aspirin plus dipyrida mole than in untreated controls.

Schondorf & Hey (1976) found no benefit fro m intravenous aspirin but had a

signi ficant reduction with aspirin combined with hep arin .

Harris, Salzman, Athanasoulis, Waltman & de Sanctis did introduce an untreated

control gro up in th eir 1977 asp irin tri al and roughl y hal ved the inc idence of phlebo-

DRUGS IN VENOUS THROMBOSIS AND PULMONARY EM BOLI SM 239

Table 5 The effectiveness of various plate let suppressant agents against deep vein thro rnbosis

(DVT) and pulmonary cmbolism (PE) in patients following hip reconstruction 11 .

N umber

01

A uthor pa tients Diagnosis Regim en PE' D V P Significance

Dechavanne 60 '211FT) Control (20 paticnts) 0 8(40%)

et al. Heparin 5000 units 0 1(5%) < 0.025

(1975) every 12h (20 patients)

Dipyridamole 150 mg 0 10(50%)

daily + aspirin 1500 mg

dail y (20 patient s)

2 Schondorf 75 12S IFT Control (15patients) 2(1) 9(60%)

&Hey Aspirin i.v. 900 mg 1 16(53%)

(1976) every 48 h (30 patients)

Aspiri n + Heparin 5000 8(27%) < 0.05

units three times daily

(30 patients)

3 Harris 95 Phlebography Control (51pat ients) I 23(45%)

et al. Aspirin 600 mg 0 11(25%) < 0.03

(1977) twicc daily (44 paticnts)

4 Rogers 73 1211FT+ Control (37 patients) 19(51%)

(1978) Phlebography

Hepari n 5000 units 0 13(36%) N.S.

twicedaily

+ sulphinpyrazone

(36 patients)

5 Silvergleid 132 12S IFT+ Control (68 patic nts) 17(25%)

et al. Phlebograph y

(1 978) Aspirin 300 mg 23(36%) N.S.

three time s dail y

+dipyridamole 75 mg

three times dail y

(64 patients)

6 Hume 71 1211FT Control (34 patients) 12(35%)

et al. Plet hys- Aspirin 600 mg 12(30%) N.S.

(1978) mograph y twice daily (37 paticnt s)

Number ofpatients developin g PE.

Number ofpatients developin g DVT.

['2Slj-ftbrinogen test.

N.S. Not significant.

graphically proven thrombi. The benefit however was only demonstrated in m a le

patients which is disappointing in the light ofthe sex d istrib ution in hip surgery.

H u me, Donaldson & Suprenant (1978), while fai ling to demonstrate a sig ni ficant

difTerence, observed a trend suggesting a favou rable resp o nse in male patients.

Roge rs et al. (19 78) combined hepa rin with sulp hi npyrazone b ut failed to dc monstrate a significan t red uction in D VT. Silverglei d , Bern stein , Burton, Tan ner,

Silverman & Sc hrier (1978) combined aspiri n a nd dipyridamole and a lso fai led to

demonstrate any reduction.

T hus aspirin a lo ne has been sho wn to be efTective in on ly o ne controlIed tria l ofhip

surgery and only in ma les . T he re rema ins uncertainty as to the optimal dose of

aspirin . Combination therapy merits furt her study. Aspirin pl us dipyridamole and

hepa rin plus sulphinpyrazone were disappointing while aspirin plus heparin showed

promise.

The advent of arachidonic acid derivatives ofTers new sco pe for prophylaxis of

venous thromboembolism - especiall y in t he va riety associated wit h vesse l wall

damage.

240 c. V. RUCKLEY

Conclusions

The role of the platelet and the interaction with the endothelial cell in thrornbogenesis is less weil defined in venous than in arterial disease.

There is no evidence from clinical studies that antiplatelet aggregating agents are

ofvalue in the treatment of established thrombosis or embolism. Prophylaxis may be

different. The different mechanisms of thrombus formation in various clinical

situations can be related to the different responses to prophylactic regimens shown by

clinical trials. A prophylactic effect has been shown with aspirin in males undergoing

elective hip surgery where vessel wall damage is a likely factor. The efficacy of other

agents such as heparin, dextrans and lignocaine, shown by clinical trial to protect

against thromboembolism may depend partlyon interference with the vessel wallblood cell interaction. The pi ace of arachidonic acid derivatives in the treatment of

venous thrombotic disease remains to be established.

References

Clagett , G. P., Brier, D. F., Rosoff, C. B., Schneider, P. B. & Saizman, E. W. (1974). Effect of

aspirin on post-operative platelet kinetics and venous thrombosis. Sur g. Forum, 25,

473-476.

Cooke , E. D., Lloyd, M. J., Bowcock, S. A. & Pilcher, M. F. (1977). Intravenous lignocaine in

prevention ofdeep venous thrombosis after elective hip surgery. Lancet, 2,797-799.

Dechavanne, M., Ville, D., Viala, J. J., Kher, A., Faivre, J., Pousset, M. B.& Dejour, H. (1975).

Controlled trial of platelet anti -aggregating agents and subcutaneous heparin in prevention

ofpost-operative deep vein thrombosis in high risk patients. Haemostasis, 4,94-100.

Fossard , D. P., Friend, J. R., Field, E. S., Corrigan , T. P., Kakkar , V. V. & Flute , P. T. (1974).

Fibrinolytic activit y and post-operative deep vein thrombosis. Lancet, 1,9-11.

French, J. E. (1965). The structure ofnatural and experimental thrombi. Ann. R oy. Coll. Surg.

Eng. , 36, 191-200.

Harris, W. H., Salzmann, E. W., Athanasoulis, C; Waltman, A. c., Baum, S. & de Sanctis,

R. W. (1974). Comparison of warfarin , low-molecular-weight dextran, aspirin and

subcutaneous heparin in prevention of venous thromboembolism following total hip

replacement. J. Bon e Joint Surg., 56, 1552-1562.

Harris , W. H., Salzman, E. W., Athanasoulis, C. A., Waltman, A. C. & dc Sanctis, R. W. (1977).

Aspirin proph ylaxis of venous thromboembolism after total hip replacement. N. Eng. J.

Med., 297, 1246-8.

Hume, M., Donaidson, W. R. & Supprenant, J. (1978). Sex, aspirin and venous thrombosis.

Orth. Clin. N. Am., 9, 761-767.

Loew, D., Brocke, P., Simma, W., Vinazzer, H., Dienst!, E. & Boehme, K. (1977).

Acetylsalicylic acid, low dose heparin and a combination of both substances in the

prevention of post-operative thromboembolism. A double blind study. Thromb. Res., 11,

81-86.

Malone , P. c., Harner , J. D. & Silver, I. A. (1979). Oxygen tension in venous valve pockets.

Abstracts VII Int Congress on Thrombosis and Haemostasis. In Thrombosis and

Haemostasis, pp. 230 . Stuttgart: Schat!auer Verlag.

Morris , G. K. & MitcheII, J. R. A. (1977). Preventing venous thromboembolism in elderly

patients with hip fractures: studies of low-dose heparin , dipyridarnole, aspirin and

flurbiprofen. Brit. med. J., 1,535-537.

MRC Steering Committee (1972). Effect of aspirin on post-operative venous thrombosis.

Lancet, 2,441-444.

Olson, P. S. (1974). Thrombus format ion in the arterial and venous circulation in hypofibrinogenemic dogs. Eur. Surg. Res., 6, 176-182 .

Olson, P. S., Lungquist , U. & Bergentz, S. E. (1974). Anal ysis of platelet, red cell and fibrin

conte nt in experimental arterial and venous thrombi. Thromb. Res., 5, 1-19.

Paterson, J. C. (1969). The pathology of venous thrombi. In Thrombosis. ed. Sherry , S., pp.

321, Washington D.C.: Nat. Acad. Sei.

DRUGS IN VENOUS THROMBOSIS AND PULMONARY EMBOLISM 241

Renney, J. T. G ., O'Sullivan, E. F. & Burke, P. E. (1976). Prevention of'post-operative deep vein

thrombosis with dipyridamole and aspirin. Brit. med. J., 1,992-994.

Rogers, P. H., Walsh, P. N., Marder, V. J., Bazak , G. c.,Lachman, J. W., Ritchie, W. G . M.,

Oppenheimer, L. & Sherry, S. (1978). Controlled trial of low-dose heparin and sulphinpyrazone to prevent venous thromboembolism after operation on the hip . J. Bone int.

Surg, 66,758-762.

Salzman, E. W. & Harris, W. H. (1976). Prevention ofvenous thromboembolism in orthopaedic

patients. J. Bone int. Surg., 58,903-913.

Salzman, E. W., Harris, W. H. & de Sanctis, R. W. (1971). Reduction in venous thrornboembolism by agents afTectingplatelet function . N. Eng. J. Med., 284, 1287-1292.

Schondorf, T. H. & Hey , D. (1976). Combined administration oflow dose heparin and aspirin as

prophylaxis ofdeep vein thrombosis after hip joint surgery. Haemostasis, 5,250-256.

Sevitt, S. (1974). Organization ofvalve pocket thrombi and the anomalies ofdouble thrombi and

valve cusp involvement. Brit. J. Surg., 61, 641-649 .

Silvergleid, A. J., Bernstein, R., Burton, D. S., Tanner, J. B., Silverman, J. F. & Schrier, S. L.

(1978). ASA-Dipyridamole prophylaxis in elective total hip replacement. Orthopedics, 1,

19-25 .

Smith, R. C; Duncanson, J., Ruckley, C. V., Webber, R. G., Allan, N. c., Dawes, J., Bolton,

A". E., Hunter, W. M., Pepper, D. S. & Cash, J. D. (1978). p-Thromboglobulin and deep

vein thrombosis. Thromb . Haemost.. 39, 338-345.

Stewart, G. J., Ritchie, W. G. M. & Lynch, P. R. (1974). Venous endothelial damage produced

by massive sticking and emigration ofleucocytes. Am. J. Path., 74,507-521.

Welch, W. G. (1887). The structure of white thrombi. Trans. Path. Soc. Philadelphia, 13,

281-285.

Wessier, S. (1955). Studies in intravascular coagulation 111. The pathogenesis ofserum induced

venous thrombosis. J. clin. Invest.. 34,647-656.

Wood, E. H., Prentice, C. R. M., McGrouther, D. A., Sinclair, J. & McNicol, G . P. (1973). Trial

of aspirin and RA233 in prevention of post-operative deep vein thrombosis. Thromb .

Diath. Haemorrh. , 30, 18-23 .

Wu, A. & Mansfield, A. O. (1980). The morphological changes of the endothelium to venous

stasis as observed under the scanning electron microscope. J. Cardiovas. Surg.. 21,

193-196.

THEVALUEOFDRUGSTHAT

MODIFY PLATELET AND VESSEL

WALL FUNCTION FOR THE

CLINICAL COMPLICATIONS

OF ATHEROSCLEROSIS

G. P. McNICOL

University Department of Medicine,

The General Infirmary.

Leeds, LSl3EX, England

The purpose of this paper is to review briefly the c1inical trial s in which an attempt

has been made to identify beneficial effects on the complications ofatherosclerosis of

drugs that are thought to modify platel et and vessel wall function. The paper is

restricted to consideration of aspirin, dipyridamole (Persantine) and sulphinpyrazone (Anturan), the only drugs with which large scale, prospective, controlled

studies have been carried out. For the purpose of this paper it will be assumed that

these drugs modify platelet and or vessel wall function in vivo in man .

Because ofits social and medical importance, and the relative ease of defming endpoints, the major c1inical model for the use of antiplatelet agents has been

myocardial infarction. At the stage of selection of drugs for testing, a number of

major problems arise. Assuming that platelets are essential factors in the genesis of

thrombosis and or atheroma, which aspects of platelet biochemistry should be the

target of therapy? Should one attempt to inh ibit format ion of pro-aggregatory

arachidonic acid derivatives, for example thromboxane A2 or prostagIandin E2, by

using for example aspirin or sulphinpyrazone which are respectively noncompetitive and competitive inhibitors of cyclo-oxygenase ? What about the relative

balance of advantage of inh ibiting thromboxane A2 formation in the platelets as

against the possibility that the same dose of drug may inhibit POh formation in the

vessel wall? Or should one attempt to elevate cyclic AMP levels by the use of

dipyridamole, a phosphodiesterase inhibitor? And then, which functional manifestations of inhibited or modified platelet biochemistry should one attempt to induce?

Should it be a prolonged platelet survival time or a prolonged bleeding time or an

impaired ristocetin-induced aggregation , or reduced platelet adhesion ex vivo to

de-endothelialised collagen ? We do not know the answers . The laboratory markers

which will predi ct c1inical benefits are not identified and there is a large element of

guess work and assumption involved in both the selection of drugs and of dosage

schedules.

The dilema is weil illustrated by aspirin; there is some evidence to suggest

DRUGS AND ATHEROSCLEROSIS 243

(O'Grady & Moncada, 1978) that a relatively small dose may be more etTective in

prolonging the bleeding time than a large dose, and still smaller doses, which might

perhaps inhibit thromboxane A2 formation in the platelets but not vessel wall PGb

production might be more etTectivethan larger doses. Again we do not know .

Primary and secondary prevention studies

Clearly on first principles the main objective in the prophylaxis of myocardial

infarction should be primary prevention. It is obviously much more attractive to prevent a patient ever having a myocardial infarction than to prevent a second or a

subsequent infarction after the initial damage has been done . Myocardial infarction is

usually associated with coronary atheroma and perhaps the logical approach to prevention of myocardial infarction is the prevention of atheroma. Is this feasible? Will

large populations of patients modify their lifestyle, for example with more exercise ,

changes in diet, the use of drugs, giving up smoking and the adoption of a relaxed

pattern ofreactions to the stresses and strains of everyday life? Even ifthese changes

can be widely implemented will they reduce the incidence of atheroma? Assuming

for the moment that the prevention of atheroma, while desirable is unlikely to be

achieved in the foreseeable future , it would seem not unreasonable to move to what

may be the next stage in the disease process, that is the inappropriate reaction of

platelets to disordered blood tlow and disordered vessel wall in atheromatous

coronary peripheral or cerebraI vessels, and to intervene at this level with drugs

which inhibit platelet function or platelet-vessel wall reactions.

Problems involved in primary prevention studies include the large number of

patients to be included in any trials, which would of necessity last many years.

Ideally high-risk patients should be identified for study , but at the present time we

lack satisfactory ways of identifying high-risk groups in the general population. The

costs involved in any such studies must be almost prohibitive.

Secondary prevention studies, that is the trial of intervention in patients who have

already declared themselves as being at high risk by a previous thrombotic episode,

for example, myocardial infarction, have the great advantage ofincluding patients in

whom a high event rate can be anticipated and hence the logistics involved in a

satisfactory trial should be much less formidable , though still large. Problems in

secondary prevention stud ies include the major difficulty that intervention is too late,

the myocardium having already been darnaged.After myocardial infarction, patients

die from many different causes , some ofwhich may be non-thrombotic in nature, for

example pump failure due to muscle death, and it is unreasonable to expect drugs

which modify platelet function or platelet vessel wall reactions to intluence this endstage process . It is also by no means clear that inhibition of platelet function would

reduce the incidence of sudden death or Iife-threatening arrhythmias after

myocardial infarction. Even in secondary prevention studies, very Iarge numbers of

patients are involved, concurrent therapy is variable and changing as other new drugs

are introduced, and major costs are involved.

The costs of secondary prevention studies include drug costs; the costs to

individual patients of haemostatic failure and other side effects; the clinicaI costs

include the work involved in prescription and supervision oftherapy; there are social

costs involved in patients' time otTwork to attend clinics, travel costs and so on , and

fmally there are the psychological costs involved in long-term ingestion of medication.

Boston Drug Surveillence Programme

Work in the field was much stimulated by the Report of the Boston Drug

244 G. P. McNICOL

Surveillence Programme (Jick & Miettinen,1976) that regular asp irin takers had only

about 20% of the risk of non-aspirin takers of leaving hospital with a discharge

diagnosis ofmyocardial infarction.

The Coronary Drug Project Research Group

The possibility that aspirin might be beneficial was reinforced by the Report of the

Coronary Drug Project Research Group (1976) in which 1629 patients admitted to

the trial, in some cases many years after myocardial infarction, there appeared to be a

benefrt from regular aspirin use (lg day"), although conventional levels ofstatistical

significance were not achieved.

Medical Research Council Trials

Two trials have been carried out by the Medical Research Council in South Wales. In

the first (Elwood, Cochrane, Burr, Sweetnam, Williams, Melsky, Hughes & Renton,

1974) 1239 men were randomlyallocated to either asp irin 300 mg daily, or placebo,

with a 12 month follow up, during which there was a 24% reduction in mortality in

the aspirin patients; this difference was not significant at the 5% level. The fmdings

were reinforced by the second Cardiff study (Elwood & Sweetnarn, 1979) in which

1682 patients were randomly allocated to aspirin 300mg three times a day, or placebo

for a year. Patients were admitted to this study early after myocardial infarction and

although there was a 17% reduction in the incidence of death in the year offollow up ,

the 5% significance level on ce again was not achieved.

Aspirin Myocardial Infarction Research Study Group (1980)

In this study, involving 4524 patients, there was a random allocation to either aspirin

(Ig day') or placebo for three years . End-points dete rmined before the beginning of

the trial included total mortality, coronary incidence (that is the combination of

coronary deaths plus proved non-fatal myocardial infarction) and fatal or non-fatal

strokes. Entry was eight weeks to five years after myocardial infarction and analysis

was by life tables and final outcome. As regards the three primary end-points, the

results for aspirin and placebo were almost identical. Not surprisingly, there was a

significant increase in gastrointestinal symptoms in patients taking aspirin.

Persantine Aspirin Reinfarction Study Research Group (1980)

This study with 2026 patients was carried out in North America and Great Britain.

There were three treatment groups; one group received 75mg dipyridamole

(Persantine) and 324 mg aspirin three times a day. The second group received aspirin

and placebo and the third group, half the size of the two previous groups, placebo

only. Treatment was for 44 months. The primary end-points were total mortality,

coronary deaths and coronary incidence. There was a trend in favour of aspirin and

dipyridamole, and of aspirin as compared with placebo, but statistical significance

was not achieved. Life table analysis however showed that dipyridamole and aspirin

were significantly better than placebo at 4, 8, 12, 16, 20 and 24 months, and aspirin

was significantly better than placebo at 8 and 24 months. Sub-group analysis showed

that the effects on the three primary end-points was much more marked in those who

entered the study within six months ofmyocardial infarction.

DRUGS AND ATHEROSCLEROSIS

Anturane Reinfarction Trial Research Group (1980)