The utilisation of aspirin as a pharmacological tool to investigate the interaction

between these two substances has been fruitful. Aspirin is active against platelet

cyclo-oxygenase in vivo and in vitro. Moreover, this elTect is long lasting because

aspirin acetylates the active site ofthe enzyme leading to irreversible inhibition (Roth

& Majerus, 1975; Roth & Siok , 1978). Platelets are unable to synthesise new protein

(Marcus, 1978) and cannot replace the cyclo-ox ygenase. Therefore, the inhibition

will only be overcome by new platelets coming into the circulation after the block of

cyclo -oxygenase in megakaryocytes has worn olT(Burch, Stanford & Majerus, 1978).

Interestingly, the cyclo-oxygenase ofvessel walls seems much less sensit ive to aspirin

than that of platelets (Baenziger, Dillender & Majerus, 1977). It has also been

suggested that endothelial cells in vitro and in vivo recover from aspirin inhibition by

regeneration ofthe cyclo-oxygenase (Czervionke, Hoak & Fry, 1978; Kelton , Hirsch ,

Carter & Buchanan, 1978). This has been reinforced by the observation that the

recovery ofthe endothelial cell synthetase in cell cultures can be pre vented by trea tment with the protein synthesis inhibitor, cycloheximide (Czervionke et al., 1979).

Studies in rabbits (Amezcua, O'Grady, Salmon & Moncada, 1978; Korbut &

Moncada, 1978) suggest that low doses of aspirin reduce TXA2formation to a greater

extent than prostacyclin formation. These experiments also showed that inhibition of

TXA2 formation is longer lasting than that of prostacyclin. Indeed, infusions of

arachidonic acid into rabbits and cats lead to an anti-thrombotic effect and to an

increase in bleeding time which can be potentiated by low doses of aspirin and

blocked by larger doses (which would inhibit prostacyclin and TXA2 formation)

(Amezcua, Parsons & Moncada, 1978; Korbut & Moncada, 1978).

Until the discovery ofprostacyclin, the use ofaspirin as an anti-thrombotic agent ,

based on its effects on platelets, looked straightforward (Majerus, 1976), although the

results of clinical trials were inconclusive (Verstraete, 1976). Now , however, the

situation needs further clarification , especially with respect to the optimal dose of

aspirin. Aspirin in high doses (200 mg kg-l) increases thrombus formation in a model

of venous thrombosis in the rabbit (Kelton et al.. 1978), and in vitro treatment of

endothelial cells with asp irin enhances thrombin-induced platelet adherence to them

(Czervionke et al.. 1978). In addition, there is an inverse correlation between platelet

adhesion and the amount ofprostacyclin produced by the tissue . Moreover, asp irin

treatment of arterial tissue in vitro increases its thrombogenicity (Baumgartner

Tschopp, 1979).

In humans, O'Grady & Moncada (1978) showed that a low single dose of asp irin

(0.3 g) increased bleeding time 2 h after ingestion, whereas a high dose (3.9 g) had no

PROSTACYCLIN, PHARMACOLOGY AND CLINICAL POTENTIAL 19

effect. Some workers have confirmed these results (Rajah, Penny & Kester, 1978),but

others have been unable to do so (Godal, Eika, Dybdahl, Daae & Larsen, 1979). The

variability might be linked to the differences in methodology or to the age of the

subjects. Indeed, Jorgensen, Olesen, Dyerberg & Stoffersen (1979) showed that the

cutaneous bleeding time in humans decreases with age and that the response to

aspirin varies according to age, being prolonged in young male volunteers but not in

older subjects. Moreover, platelet aggregation and TXA2 formation are blocked 2 h

after a single high dose of aspirin (3.9 g). The bleeding time is unchanged at that time

but 24 and 72 h after aspirin it is increased and slowly recovers towards pretreatment

levels over aperiod of 168 h, in a manner which mirrors the recovery of TXA2

formation and platelet aggregability (Amezcua et al., 1979). An extension of the

concept comes from the demonstration that tranylcypromine, an inhibitor ofprostacyclin formation, enhances platelet aggregation in an experimental model of thrombosis in the microcirculation of the brain of the mouse (Rosenblum & EI-Sabban,

1978). All these results clearly demonstrate that the prostacyclin /thromboxane

balance is an important mechanism of control of platelet aggregability in vivo.

Clearly, manipulation of this control mechanism might lead to pro- or antithrombotic states of clinical relevance. In this context it is interesting that hydrocortisone treatment of normal or thrombocytopenic rats blocks prostacyclin formation in the vessel wall and decreases the bleeding time (Blajchman, Senyi, Hirsh,

Surya, Buchanan & Mustard, 1979),a result which would be expected from the interference with arachidonic acid release induced by steroids (Flower, 1978).The authors

(Blajchman et al., 1979)mention that for years it has been the clinical impression that

steroids decrease the bleeding time in thrombocytopenic patients without increasing

the platelet count.

Studies in man attempting to measure TXB2 and (or) prostacyclin or 6-oxoPGFw after different aspirin dose schedules have confirmed the higher sensitivity of

platelet cyclo-oxygenase to aspirin. Masotti, Galanti, Poggesi,Abbate & Neri Serneri

(1979)found that aspirin at 3.2-3.4 mg kg-I gave a 50% inhibition of ex vivo platelet

aggregation by several agents, while 4.9 mg kg-I was needed for 50% inhibition of

prostacyclin formation. It has also recently been demonstrated that a single daily

dose of aspirin (160 mg) protects against thrombosis in arterio-venous shunts in

patients. This dose reduced significantly (40%) the incidence of thrombosis over a

five month observation period (Harter, Burch, Majerus, Standford, Pelmes, Anderson & Weerts, 1979).

From all these results it is clear that a selective inhibitor ofthromboxane formation

should now be tested for antithrombotic efficacy, since theoretically this provides an

advantage over aspirin in allowing prostacyclin formation by vessel walls or other

cells either from their own endoperoxides or from those released by platelets. This

should be the main criteria for determining a 'superior' mechanism of action over a

small dose of aspirin . Studies in vivo are not yet available but in vitro Needleman,

Wyche & Raz (1979) made the observation that when platelets were treated with a

TXA2 synthetase inhibitor, then endoperoxides became available for utilisation by

the vessel wall. Interestingly, in the presence of a thromboxane synthetase inhibitor,

arachidonic acid or collagen added to blood in vitro lead to the formation of 6-oxoPGFla rather than TXB2, showing that some cells other than platelets have

synthesised prostacyclin (Blackwell et al., 1978).These results support the suggestion

that thromboxane synthetase inhibitors might have a superior antithrombotic effect

to cyclo-oxygenase inhibitors (Moncada & Vane, 1977, 1978). It is important to

realize at this stage, however, that all these observations have been made in vitro and

that in vivo experiments are necessary in order to clarify further the nature of the

interaction between platelets and normal or damaged vessel walls.

Whether other drugs exert their antithrombotic effect by acting on the prostacyclin/thromboxane system is not yet known but studies using sulphinpyrazone in

20 S. MONCADA & J. R. VANE

cultured endothelial cells (Gordon & Pearson, 1978) and ticlopidine given orally to

rats (Ashida & Abiko, 1978) suggeststhat these compounds have little or no efTect on

prostacyclin formation at concentrations at which they afTect platelet behaviour. A

compound which might stimulate prostacyclin formation in humans after oral

ingestion has also been described (Vermylen, Chamone & Verstraete, 1979).

Prostacyclin and the cardiovascular system

Prostacyclin relaxes most vascular strips in vitro including rabbit coeliac and

mesenteric arteries (Bunting , Gryglewski, Moncada & Vane, 1976a), bovine coronary

arteries (Dusting, Moncada & Vane , 1977b; Needleman, Bronson , Wyche, SivakofT&

Nicolaou, 1978), human and baboon cerebral arteries (Boullin, Bunting, Blaso, Hunt

& Moncada, 1979) and lamb ductus arteriosus (Coceani , Bishai, White, Bodach &

Olley, 1978). Exceptions to this include the porcine coronary arteries (Dusting,

Moncada & Vane , 1977c) and some strips of rat venous tissue and isolated human

saphenous vein (Levy, 1978) which are weakly contracted by prostacyclin. Whether

these constrictor efTects are induced in the corresponding circulations in the intact

animal or man has not been studied. In strips of human umbilical artery, prostacyclin induces a dose-dependent relaxation at low concentrations « IO-6M) and a

dose-dependent contraction at higher concentrations (> 1O-5M) (Pomerantz, Sintetos

& Ramwell, 1978). As mentioned earlier, prostacyclin and not PGE2 is the main

metabolite of arachidonic acid in isolated vascular tissue, and this has led to an

intense study to reassess the efTects and role of arach idonic acid and its metabolites in

vascular tissue and the cardiovascular system.

Unlike other prostaglandins such as PGE2 and PGF2a, prostacyclin is not

inactivated in vivo on passage through the pulmonary circulation. Indeed, the lungs

constantly release small amounts ofprostacyclin into the passing blood (Gryglewski,

Korbut & Ocetkiewicz, 1978b; Moncada, Korbut, Bunting & Vane , 1978) perhaps

because of the huge massof endothelial cells present. The concentration of

prostacyclin is higher in arterial than in venous blood for there is about 50% overall

inactivation in one circulation through peripheral tissues (Dusting et al., 1977a;

Dusting, Moncada & Vane, 1978c). Recent work in humans measuring 6-oxO-PGFIU

in the arterial and venous side of the circulation during cardiac catheterisation

confirms these difTerences (Hensby, Barnes, Dollery & Dargie, 1979).

Platelets, therefore, may be constantly influenced by circulating prostacyclin and

consequently they can have higher cyclic AMP levels and be less aggregable than has

ever been detected by in vitro measurements which are only made after a 10-30 min

delay during which the blood is processed. In this period, prostacyclin and its efTects

will decay. This concept explains the difTerences in reactivity between platelets in

vitro and in vivo reported by many authors and might throw light on recent

controversies about control ofcyclic AMP levels in platelets.

In the anaesthetised dog, prostacyclin is hypotensive in doses ranging from 50-1000

kg-Imin-t (Armstrong, Chapple, Dusting, Hughes, Moncada & Vane , 1977). Intravenously in anaesthetized rabbit or rat, prostacyclin causes a fall in blood pressure

and is 4-8 times more potent that PGE2. Prostacyclin is at least 100 times more active

than its degradation product, 6-oxo-PGF.u (Armstrong et al., 1977). Since it is not

inactivated by the pulmonary circulation, prostacyclin is equipotent as a vasodilator

when given either intra-arterially or intravenously in the rat , rabbit or dog (Armstrong, Lattimer, Moncada & Vane, 1978; Dusting et al.,1977a; Dusting et al., 1978c).

This is an importarit difTerence from PGE. or PGE2 which, because of strong

pulmonary metabolism, are much less active when given intravenously (Ferreira &

Vane , 1967).

PROSTACYCLl N, PHARMA COLOGY AND CLINICAL POT ENTIAL 21

In the heart , local injections of arachidonic acid into the coronary circulation of

the dog cause vasodilatation, and because this effect is abolished by indomethacin

(Hintze & Kaley, 1977) it was assumed that PGE2was the likely mediator. However ,

there were some major difficulties with this proposal. In isolated Langendorffperfused hearts of the rabbit, arachidonic acid dilated the coronary vasculature, but

PGE 2 was inactive (Block, Feinberg, Herbaczynska-Cedro & Vane , 1975; Needleman , 1976).Isolated strips ofbovine, canine and human coronary artery were relaxed

by arachidonic acid but PGE2 contracted them (Kulkarni, Roberts & Needleman,

1976). Arachidonate-induced relaxation of these strips was abolished by

indomethacin and it was suggested, therefore, that the metabolite responsible must be

the endoperoxide intermediate PGH2 (Kulkarni eral., 1976).

Later, Dusting eral., (l977b) showed that bovine coronary arteries were relaxed by

prostacyclin and PGH2 (which sometimes induced an initial transient contraction),

and after treatment with 15-HPAA (an inhibitor of prostacyclin synthetase) the

relaxation induced by arachidonic acid was abolished, whilst that induced by PGH2

was reversed to a contraction. Thus, relaxation of coronary arteries induced by

arachidonic acid or PGH2 is due to intramural metabolism to prostacyclin. This

study further confirmed that the intrinsic activity ofPGH2 on isolated blood vessels is

contractile (Dusting eral., 1977b). Similar results have been published (Needleman er

al.. 1978; Raz, Isakson, Minkes & Needleman, 1977).

In isolated Langendorff-perfused hearts of the guinea pig and rabbit , not only is

prostacyclin a potent vasodilator but it is also the predominant metabolite of arachidonic acid (Schror, Moncada, Ubatuba & Vane , 1978). Similarl y, others have

identified 6-oxo-PGFI a as the rnajor product from rat and rabbit hearts perfused

with arachidonic acid (De Dekere , Nugteren & Ten Hoor , 1977). The coronary

actions of prostacyclin in the intact heart of open ehest dogs have been examined

(Armstrong er al., 1977; Dusting, Chapple, Hughes , Moncada & Vane, 1978a;

Hyman, Kadowitz, Lands, Crawford , Fried & Barton , 1978). Local injection of

prostacyclin (50-500 ng) into the coronary circulation increased coronary blood flow

without systemic effects and it was a more potent coronary dilator than PGE2.

Furthermore, profound and prolonged coronary vasodilatation was rapidly elicited

by prostacycl in (20-100 1Jg) absorbed through the myocardium after dripping a

solution on to the surface ofthe left ventricle (Dusting eral., 1978a). Interestingly, the

coronary circulation is sensitised to the vasodilator effects of exogenous prostacyclin,

but not to those ofPGE2, when endogenous synthesis is inhibited by indomethacin or

meclofenamate (Dusting er al., 1978a; Hintze & Kaley, 1977). These inhibitors of

cyclo-oxygenase decrease resting coronary blood flow in anaesthetised, open ehest

dogs. Although this is not seen in conscious dogs without acute surgery (Owen,

Ehrhart, Weidner, Scott & Hadd y, 1975), it does indicate that the generation of a

vasodilator metabolite of arachidonic acid increases or maintains coronary blood

flow during mildly traumatic conditions. It is clear that this metabolite is

prostacyclin.

Bradycardia accompanying the hypotension induced by prostacyclin has been

observed in anaesthetised dogs (Armstrong eral., 1977; Dusting eral., 1978a; Hintze,

Kaley, Martin & Messina, 1978) and only transient weak tachycardia accompanied

prostacyclin infusion in anaesthetised cats (Lefer, Ogletree , Smith , Silver, Nicolaou,

Barnette & Gasic, 1978). Bradycardia induced by prostacycl in is a reflex response

mediated at least partially by vagal pathways since atropine reduces or abolishes the

bradycardia (Chapple, Dusting , Hughes & Vane, 1978, 1980). However, the afferent

arc is also subserved by vagal fibres, for vagotom y (but not atropine treatment)

reduces the hypotensive effects of prostacyclin. Therefore, the hypotension induced

by prostacyclin has at least two components: direct arteriolar vasodilatation and

reflex, non-cholinergic vasodilatation. Similar results have been obtained by Hintze

eral.(1978).

22 S. MONCADA & J. R. VANE

In the dog, prostacyclin infused intravenously at rates below those needed for a

systemic effect reduces renal vascular resistance and increases renal blood flow and

urinary excretion ofsodium, potassium and chloride ions (Bolger, Eisner, Ramwell &

Slotkoff, 1978; HilI & Moncada, 1979). There is increasing evidence that prostacyclin

mediates the release of renin from the renal cortex . Arachidonic acid, prostaglandin

endoperoxides or prostacyclin all stimulate renin release from slices of rabbit renal

cortex, but PGE2 has no such effect (Weber, Larsson , Anggard, Hamberg, Corey ,

Nicolaou & Samuelsson, 1976; Whorton, Misono, Hollifield, Frolich, Inagami &

Oates, 1977a). Furthermore, indomethacin reduces renin release in animals and man

(Data, Crump, Hollifleld, Frolich & Nies, 1976; Frolich, HollifIeld, Dormois,

Frolich, Seyberth, Michelakis & Oates , 1976; Larsson , Weber & Anggard, 1974).

Prostacyclin-like activity and 6-oxo-PGFlu have been identified in ineubates of

PGG2 or PGH2 with renal cortical microsomes (Remuzzi, Cavenaghi, Mecca, Donati

& De Gaetano, 1978a; Whorton, Smigel, Oates & Frolieh, 1977b; Zenser, Herman,

Gorman & Davis, 1977). Thus, prostacyclin may be the obligatory endogenous

mediator of renin secretion by the kidney. Indeed, Gerber, Braneh , Nies, Gerkens,

Shand, Hollifield & Oates (1978) have demonstrated that prostacyclin induces renin

release when infused intrarenally into dogs, and HilI, Moneada & Vane (1978) have

demonstrated increased concentrations of angiotensin 1I in arterial blood during

intrarenal infusions of prostacyclin. 6-oxo-PGFIU is also formed by colleeting tubule

cells isolated from rabbit papillae (Grenier & Smith , 1978). Interestingly, angiotensin

11 releases prostaeyclin from the rat kidney in vitro(Silberbauer, Sinzinger & Winter,

1979)and the dog kidney in vivo(Mullane, Moncada & Vane, 1979b).

Prostacyclin is also a strong vasodilator in the mesenteric and hind limb cireulations ofthe dog (where TXA2 is a vasoconstrictor) (Dusting eral.. 1978b) and on the

precapilIary side of the microcirculation of the hamster eheek pouch (Higgs,

Cardinal, Moncada & Vane, 1979), where it also reverses eateeholamine-induced

vasoconstriction. In this preparation ö-oxo-PGf'm had l/20th the vasodilator

aetivity of prostaeyclin and was more potent than PGE2. In the pulmonary eireulation ofthe dog, prostacyclin is the only product ofarachidonic acid which produces

strong vasodilatation (Kadowitz, Chapnick, Feigen, Hyman, Nelson & Spannhake,

1978; Mullane, Dusting, Salmon, Moncada & Vane, 1979a). It also dilates the

pulmonary vaseular bed of the foetal lamb where its potency is greater than PGEI

but less than PGE 2(Leffier & Hessler, 1979).

Prostacyclin in man

Prostacyclin has potent effects on platelets and on the cardiovascular system in man ,

as first demonstrated by Gr yglewski, Szczeklik & Nizankowski (l978d) and

Szczeklik, Gr yglewski, Nizankow ska, Nizankowski & Musial (l978b) .

During infusion of prostaeyclin in healthy volunteers for 60 min doses ranging

from 0.5-16 ng kg-I min-I there was a dose related inhibition of platelet aggregation

measured in platelet rieh plasma and in whole blood at doses of 2-16 ng kg-t min-I

(O'Grady, Warrington, Moti , Bunting, Flower, Fowle, Higgs & Moncada, 1979).

Similar inhibition of platelet aggregation was seen when the responses were measured

at 15 or 45 min after start of the infusion . At a dose of 8 ng kg! min-I partial

inhibition of aggregation was demonstrable for up to 105 min after the end of

infusion and this persistence of effect on platelet has recently been confirmed

(Chierchia, Ciabattoni, Cinotti, Maseri, Patrono, Pulgiese, Distante, Simonetti &

Bernini, 1979). Template bleeding time was not significantly increased although

Szczeklik eral. (l978b) found an approximate doubling ofbleeding time in response

to prostaeyclin at 20 ng kg-I min-1•

PROSTACYCLIN, PHARMACOLOGY AND CLINICAL POTENTIAL 23

Prostacyclin disperses circulating platelet aggregates (Szczeklik er al.. 1979b).

Significant changes in the response curve of platelet aggregation to log dose ADP

were seen (Fitzgerald, Friedman, Miyamori, O'Grady & Lewis, 1979) when prostacyclin was administered under double-blind conditions at rates of 4 and 8 ng kg-1

min-I.

Other haematological variables such as platelet count, platelet factor 3 concentration , accelerated partial thromboplastin time, prothrombin time, euglobin clot lysis

time, concentration of fibrinogen degradation products and blood glucose were not

affected by prostacyclin (O'Grady eral.. 1979;Szczeklik eral.. 1979b).

It was originally suggested (Szczeklik et al., 1978b) that prostacyclin had direct

positive chronotropic and inotropic effects in man . However, in a double-blind controlled study using prostacyclin at 0-4 ng kg-I mirr! an increase in heart rate accornpanied by decrease in diastolic blood pressure, pre-ejection period and QS2 index was

observed (Warrington & O'Grady, submitted for publication). Systolic blood pressure, left ventricular ejection time index and the normalised first derivative of the

apex cardiogram were unaltered by prostacyclin. These findings were consistent with

an arteriolar vasodilator effect of prostacyclin which would be expected to lower

diastolic and mean blood pressure and thus reflexly increase heart rate and contractility.

When heart rate was increased by more than 10% over control values during

prostacyclin infusion, peripheral temperature at the great toe increased by 1-6"C

(O'Grady er al., 1979), Increases in skin temperature as weil as facial flushing were

also observed at doses of 2-5 ng kg-I mirr'! (Szczeklik er al.• 1979b). Facial flushing

invariably occurs at doses above 4 ng kg-1 min-t when an increase in heart rate of

more than 10% is recorded (O'Grady eral.. 1979). This flushing limits the extent to

which studies with prostacyclin can be rendered double-blind. The cardiovascular

effects are shorter-lived than those on platelets and disappear within 15 min of the

end ofinfusion (O'Grady et al., 1979).

Plasma renin activity rises significantly during prostacyclin infusion (Fitzgerald er

al., 1979). In the same study plasma noradrenaline and plasma aldosterone levels did

not change significantly, Renal blood flow measured using (I25I]-hippuran increased

in response to a dose of prostacyclin (6 ng kg! min- 1

) which caused a small reduction

in diastolic blood pressure while the glomerular filtration rate measured using [SICr]-

EDT A remained unchanged (Henry & O'Grady, unpublished results).

Headache has been reported by many subjects when infusion rates greater than 8

ng kg-1 mirr" are used (Fitzgerald er al., 1979; O'Grady et al.• 1979; Szczeklik er al.,

1978b), Colicky central abdominal discomfort has been less frequently experienced

but was reproducible in one subject (O'Grady et al., 1979), The precise mechanism of

these gastrointestinal effects is unclear. It may be that they reflect contraction of

human gastrointestinal smooth muscle by prostacyclin; they mayaiso be vagally

mediated or represent secondary effectsof prostacyclin or of its metabolic products.

III defmed sensations of unease and restlessness have been experienced by subjects

receiving higher doses of prostacyclin (Chierchia et al.. 1979; O'Grady er al.. 1979;

Szczeklik er al.. 1978b). At the high dose of prostacyclin of 50 ng kg-1 min- 1

administered to two subjects (Szczeklik et al.. 1978b) both experienced sudden weakness with pallor and nausea, systolic and diastolic blood pressure fell and bradycardia

occurred . It is possible that this effect is mediated by a vagal reflex, for in dogs

prostacyclin produces a vagally dependent bradycardia (Chapple et al.. 1978).

Following reports that PGEt has been used successfully in the treatment of

peripheral vascular disease (Carlson & Olsson, 1976) prostacyclin has been shown to

have a similar effect, producing a long lasting increase in muscle blood flow, disappearance of ischaemic pain and healing of trophic ulcers after an intra-arterial

infusion to the affected limb for three days (Szczeklik, Nizankowski, Skawinski,

Szczeklik, Gluszko &Gryglewski, 1979).

24 S. MONCADA & J. R. VANE

The circulation of blood through extracorporeal systems involves the blood

coming into contact with an artificial surface which is unable to generate prostacyclin. In the course of such procedures thrombocytopenia and loss of platelet

haemostatic function occur and make an important contribution to the bleeding

problems following charcoal haemoperfusion and prolonged cardiopulmonary

bypass in man (Friedenberg, Myers, Plotka, Beathard, Kummer, Gatlin, Stoiber, Ray

& Sautter, 1978; Moriau, Masure, Hurlet, Debeys, Chalant, Poulot, Jaumain,

Servaye-Kestens, Ravaux, Louis & Goenen, 1977; Weston, Rubin, Hanid, Langley,

Westaby & Williams, 1977). Formation of microemboli during cardiopulmonary

bypass mayaiso contribute to cerebral complications which sometimes follow this

procedure (Patterson & Kessler, 1969). In animals subjected to experimental renal

dialysis (Woods, Ash, Weston, Bunting, Moncada & Vane, 1978), charcoal haemoperfusion (Bunting, Moncada, Vane, Woods & Weston, 1979)and cardiopulmonary

bypass (Longmore, Bennett, Gueirrara, Smith, Bunting, Moncada, Reed, Read &

Vane, 1979) infusion of prostacyclin during the procedure prevented this platelet

damage and thrombocytopenia thus increasing the biocompatability of the proöedure. These fmdings have been confirmed in patients with fulminant hepatic

failure undergoig charcoal haemoperfusion (Gimson, Hughes, Mellon, Woods,

Langley, Canalese, Williams & Weston, 1980). Prostacyclin infusion prevented the

fall in platelet count and elevation of ßthromboglobulin seen in the control patients.

In addition two of the control patients developed marked hypotension during the

procedure, in one associated with a marked rise in Swank Screen filtration pressure,

while this did not occur in the prostacyclin treated patients. A study of serial

haemoperfusion with prostacyclin on the survival rate of patients with fulminant

hepatic failure is now in progress.

During cardiopulmonary bypass in man preliminary indications (Bunting,

O'Grady, Moncada , Vane, Fabiani , Terrier & Dubost, unpublished results 1980)are

that in patients receiving prostacyclin during bypass platelet number and function

are better preserved and marked rises in Swank Screen filtration pressure prevented.

During the course ofstudies on extracorporeal circulation systems it was observed

that prostacyclin potentiates the effects of heparin (Bunting et al., 1979). Further

studies on this interaction demonstrated that prostacyclin has also a small indirect

anticoagulant effect. Indeed platelets stimulated by low doses of aggregating agents

accelerate clotting by providing a surface upon which coagulation factors can

combine and react more efficiently (see Marcus, 1978). Prostacyclin, by preventing

platelet activation, inhibits the shortening of clotting time produced when either

kaolin or collagen are incubated with platelet rich plasma (Bunting & Moncada,

1980). Platelets release anti heparin activity which in vitro reduces the anticoagulant

effect of heparin . Prostacyclin by inhibiting this release and by preventing the

development of pro-coagulant activity can potentiate the action of heparin as much

as one hundred per cent (Bunting & Moncada, 1980). These in vitro fmdings agree

with the observations in extracorporeal circulations(Bunting et al., 1979).

Potentialforthedevelopment of antithrombotic therapy

Intra-arterial thrombus formation and haemostatic plug formation have been

described in general terms as equivalent phenomena (Mustard & Packham , 1975). It

is, however, possible that the relative importance of prostacyclin and TXAz in both

conditions is different, for prostacyclin is an unstable circulating hormone

(Gryglewski et al.. 1978b; Moncada et al., 1978)as weIlas a locally generated one. Its

role in controlling intra-arterial thrombus formation might be more important than

that of TXAz which seems to be synthesised only after strong interaction with

collagenous structures by aggregatingplatelets. The opposite situation might be true

PROSTACYCLIN , PHARMACOLOGY ANDCLINICAL POTENTIAL 25

during haemostatic plug formation when platelets are out ofthe vessellumen and in

strong interaction with pro-aggregating structures ofthe vessel wall and surrounding

tissue. In these conditions TXAI formation might be predominant. Ifthe PGh/TXAI

system in general is considered as a defensive homeostatic mechanism where platelet

aggregation is undesirable inside the vasculature but is necessary for the arrest of

bleeding, then such a balance would be expedient in biological terms.