Archives of Disease in Childhood, 1980, 55, 99-105

Annotations

Patent ductus arteriosus: experimental aspects

In the fetus the ductus arteriosus joins the pulmonarytrunk to the aorta and is of a calibre similar tothese major vessels. Its structure is different sinceelastic tissue is normally absent from the muscularlayers, and its physiological behaviour differsquantitatively and qualitatively from that of thevessels it connects.

In fetal life the ductus provides a bypass to thelungs such that (in the lamb) almost nine-tenths oftheright ventricular output is directed to the lower bodyand placenta.' When the lungs expand with gas forthe first time, pulmonary arterial pressure falls fromslightly above to substantially below aortic level andflow through the ductus reverses.2 This, pendingconstriction and final closure of the ductus,inaugurates a left-to-right shunt recirculating aproportion of left ventricular output through thelungs. As failure of the ductus to close normally isassociated with cardiac embarrassment, it isimportant to understand the mechanisms whichpromote closure, why they do not normally operatein utero, and whether it is possible to assist closurewithout surgery. Conversely, the adverse postnataleffects of some congenital malformations in whichblood can only reach the lungs in any quantity via theductus are alleviated temporarily by treatment whichmaintains ductal patency pending corrective surgery.The further possibility exists that substances capableof crossing the placenta can, when administered tothe mother, affect the ductus and modify the distri-bution of fetal cardiac output.

This note is concerned not with ultimate oblitera-tion of the ductus, but with the constriction whichtakes place shortly after birth.The fact that initial constriction of the ductus

occurs abruptly was first realised nearly 80 years ago,although appreciation of the anatomical idio-syncracies of the fetal circulation is of course mucholder. Experimental investigation of the mechanismsof ductus closure began about 40 years ago whenBarcroft et al.3 and Kennedy and Clark4 obtainedresults on guinea-pigs which suggested that oxygenwas implicated.More extensive experiments on fetal lambs

showed that increase in °2 content of the bloodconsistently constricted the ductus even afterdestruction of the brain and spinal cord and when

the pulmonary and systemic arterial pressures werestabilised.2 The response was also obtained in anisolated heart-ductus-artificial lung preparation andwas reversed by anoxia.

In 1963 Kovalcik5 confirmed in vitro the directcontractile action of oxygen on ductus muscleisolated from guinea-pigs and lambs, and showedthat calcium was essential for its occurrence. Sincethen more penetrating investigation of the oxidativemetabolism of ductus muscle has been undertaken6-8and an extensive range of smooth muscle stimulantsinvestigated. These have included such diversesubstances as potassium and acetylcholine which, inappropriate concentration, regularly, or in the case ofbradykinin less regularly, constrict ductus muscle.>6 9They remain useful as experimental tools but seemunlikely participants in natural constriction of theductus. Since 1972 interest has turned to the partprostaglandins play in the behaviour of this remark-able vessel.

Isolated ductus muscle

Preparations of ductus muscle have been extensivelyinvestigated in the form of spiral strips or rings andas the isolated perfused vessel in vitro or in vivo.

Oxygen. Fay6 found that the maximum tensionevoked in ductus muscle by an increase of Po2 wasdouble that produced by acetylcholine or potassium,which stimulate the muscle directly and elicitmaintained contraction even under anaerobicconditions. The ductus was equally sensitive tochange of luminal or adventitial Po2, and half-maximal contraction was attained with a Po2 of70 mmHg (9-31 kPa) both inside and outside. Thismakes unlikely the involvement of specialised02-sensitive cells. Inhibitors of oxidative phos-phorylation reduced the contractile response to 02to one-third or less, but had less effect on acetyl-choline-induced contractions.While Kovalcik5 had earlier found that the

phenothiazines and amytal prevented the ductus fromcontracting in response to °2, in his hands sodiumcyanide 1 mmol/l only had a limited effect and hetherefore doubted whether the terminal steps of the

99

100 Joan C Mott

electron transport chain were concerned. Hesuggested a flavoprotein oxidase not inhibited bycyanide might be involved. Under the conditions ofhis experiments the responses were often slow, and itis possible that cyanide did not remain stable for thelength of time required.Fay and J6bsis7 demonstrated that carbon

monoxide also depressed 02-induced contraction;this effect was reversible by light. Dissociation of theCO-cytochrome a. complex preceded muscularcontraction by 81 ± 13 (SEM) seconds, thusestablishing that the temporal sequence of events iscompatible with the concept that 02 acts via thecytochrome respiratory chain. They found that both02 consumption (which was relatively high forsmooth muscle) and muscle tension were similarlyrelated to 02 tension. The 03 consumption of aductus partly constricted by 02 remained unchangedwhen further tension was elicited by the addition ofacetylcholine. This suggested that there is indeed alimitation on cellular respiration imposed by 02availability.Fay6 calculated that only when external Po2

exceeded 80 mmHg (1064 kPa) would all the cellsof the ductus smooth muscle be likely to receivesufficient 02 to sustain their maximum respiratoryrate. This would be compatible with a Km 02 ofabout 1 mmHg for the terminal cytochrome a3.Direct measurement showed very steep oxygengradients in the muscle.8

02-evoked contractions of ductus muscle weredistinguished by stepwise increases of tension neverseen with acetylcholine and other direct stimulants.Fay68 likened these to those of other smooth musclessuch as taeniae coli which exhibit propagated actionpotentials, and suggested that these might be fuelledby increased synthesis of high-energy phosphatecompounds via the respiratory chain.Clyman et al. found that the maximum tension of

02-induced contraction of lamb ductus ringsincreased with increasing gestational age, and alsothat ductus tissue from immature lambs was verysensitive to light.10 They thought that this photo-relaxation might account for the relative insensitivityof the immature ductus to stimulation by 02 in vitro.It is interesting that the response of the muscle todirect stimulation with potassium does not varywith age in lambs or guinea-pigs.9

Prostaglandins. Clarification of the intricate relation-ships and actions of these naturally occurring lipidderivatives had proceeded far enough by 1972 forCoceani and Olley"l-12 to examine their action onisolated strips of lamb ductus arteriosus. Contrary toexpectation, they observed that PGE1 and PGE2relaxed the anaerobic muscle but had less action on

the aerobic (contracted) muscle. The threshold dosewas 10-9 mol/l; PGA, and PGF,, also causedrelaxation but only at higher concentrations. Coceaniand Olley12 suggested that the ductus might bemaintained open in fetal life by PGE compoundswhich also relax the calf ductus. At about the sametime, Elliott and Starling13 reported that PGF2Xcontracted calf ductus arteriosus (though'at a highdose 10-4 mol/l) as Coceani and Olley12 hadenvisaged for this class ofcompounds in general.

Starling and Elliott14 gave a systematic account ofthe responses of the calf ductus to prostaglandins(3-3 x 10-5 mol/l) at 03 tensions between 15 and100 mmHg (2 0 and 13 3 kPa). PGF2Q was alwaysconstrictor and acted synergistically with 0,2 PGEcaused relaxation at all levels of 0°.PGE, initially caused some contraction followed

by relaxation, most pronounced at lower 02 tensions.They suggested that PGF2, might be responsible forclosure of the ductus in calves, and considered thatthe inhibition of contraction by indomethacin andnaproxen implied endogenous synthesis ofprostaglandin. In the few experiments attempted, thehuman ductus behaved in a manner similar to thecalf's.

Coceani et al.'5 examined the effects of inhibitorsof PGE2 synthesis on the response of lamb ductusarteriosus to PGE2 at both low and high 02 levels.The relaxant effect of PGE2 was reduced at high 03levels, but could be restored after treatment with PGsynthetase inhibitors. Possibly the receptiveness ofthe target site is controlled by the rate ofendogenousPG formation. They suggested that at birth PGEformation may increase but the tissue become lesssensitive. Olley et al.16 found that ibuprofen (anotherprostaglandin synthetase inhibitor) caused contrac-tion of the ductus from lambs as young as 90 days'gestation.Clyman et al.17 did not encounter the insensitivity

of well oxygenated lamb ductus muscle to therelaxant action of PGE reported by Starling andElliott14 in calf ductus, and suggested on the basis oftheir own carefully designed investigation that theearlier experiments might have been vitiated bytachyphyllaxis. However, Clyman et al.'8 routinelyused much higher 02 pressures than Starling andElliott. Indomethacin-induced contraction of lambductus was, as the results of Coceani et al.'5 hadindicated, additive to 02-induced contraction,18 asituation consistent with endogenous prostaglandinsynthesis. In lambs less than 110 days' gestational ageindomethacin-induced contraction is greater and theED2, for relaxation by PGE2 smaller than in lambsnear term.19

Coceani et al.0 and Clyman et al. 21 have examinedthe action of several prostaglandins and some of their

Patent ductus arteriosus: experimental aspects 101

derivatives on the lamb ductus. Only PGF2, and its1 5-ketometabolites have been found to causecontraction. Those causing relaxation and effectivebelow 10-8 mol/l were PGE1, PGE2, and their 13, 14dihydrometabolites. The threshold for PGE1 is10-12 mol/l and for PGE2 10-1" to 10-12. Both causecontraction above 10-6 and 5 x 10-6 mol/l re-spectively.The isolated perfused ductus from guinea-pigs and

rabbits constricts in response to PGF2, but dilateswith PGE, and PGE2.22 Dog pup ductus also dilateswith PGE.22 However the contractile action ofprostaglandin synthetase inhibitors on the guinea-pigductus is small.23Smiesko et al.24 found the ductus from fetal

guinea-pigs and rabbits maintained in nitrogenconstricted when the transmural pressure wassuddenly increased. There was a concurrent reductionof sensitivity to vasoconstrictors, and it was suggestedthat this might be implicated in failure of the ductusto close. Whether the initial myogenic responseobserved was attributable to prostaglandin releaseafter mechanical stimulation is a matter for specula-tion.

Prostaglandin synthesis. Homogenates of term ductusof lambs2.26 and calves27 converted exogenousarachidonic acid to prostaglandins E2, F2,, and6-keto-F1,. The major product was 6-keto-F,2which has only weak vasoactive properties, but itsshort-lived precursor (a prostacyclin) PGI2 is a weakvasodilator. More PGI2 was produced by ductustissue from immature lambs than mature ones.26Although only about one-tenth of the prostaglandinsynthesised is PGE2, ductus muscle is so muchmore sensitive (threshold 10-1" mol/l) to it than toPGI2 (10-8 mol/l), that it seems likely to be themajor prostaglandin maintaining relaxation of theductus.

Cultures of vascular endothelium, especially of theumbilical vein, can synthesise PG12 and PGE.Angiotensin II 10-7 mol/l stimulates production ofPGE in such a system.28 Bradykinin has also beenreported to increase vascular synthesis of prosta-glandins.29 The endothelium of the ductus has notreceived comparable attention in relation to thesecompounds.

It is interesting that reduced glutathione whichfavours PGE synthesis has been found to relax thehypoxic ductus.30

The whole animal

Anatomical investigations. Sharpe and Larsson usedwhole body freezing techniques31 to examine the

effects of prostaglandin administration on the ductusof newborn rats and rabbits. They observed that innewborn rats killed 30 minutes after subcutaneousinjection of PGF2,, the ductus diameter was aboutfour times that found in saline-treated controls. Theresponse was dose-dependent over the range 0-2 to5 Fg/g body weight and subsided by 60 minutes afterinjection. PGE1 0 - 2 Vg/g caused dilatation compar-able with that seen with PGF2, (1 ,ug/g) and similarresults were obtained with both prostaglandins innewborn rabbits. In the absence of hypoxaemia andacidosis it was concluded that both PGF2L andPGE1 had delayed ductal closure.

Starling et al.32 obtained radiographical evidencein normoxic piglets 3-6 hours of age of dilatation ofthe ductus in response to infusion of PGE1 or PGE2(1-4 ug/kg per min) or PGA1 or PGA2 (20-40 ,tg/kgper min) into the aortic arch; the left to right shuntappeared larger.The obstetric use of the prostaglandin synthetase

inhibitors, indomethacin and acetylsalicylic acid, inconjunction with the fact that the concentration ofindomethacin in human fetal blood can reach 1-6times that in the mother33 prompted Sharpe et al.34-35to investigate the fetal consequences of maternaladministration of indomethacin 15 mg/kg. Thiscaused constriction of the ductus in near term fetalrabbits and rats. Rat pups so exposed in utero werecyanosed and lethargic on delivery with distressedbreathing. Doses of indomethacin (2 * 5 mg/kg) in theclinical range and acetylsalicylic acid (50-100mg/kg) were also effective.Indomethacin 2 x 5 mg/kg per day was given

orally to 10 near term ewes for 3 days with the finaldose 1-3 hours before caesarean section.36 The lambswhose blood-gases were in the normal range weredelivered into an hypoxic atmosphere (<14 mmHg1 9 kPa), the cord was tied and the chest openedbefore immersion in liquid nitrogen. The cross-sectional areas of the ductus ranged from 0 * 84-11 *03mm2 and were less (P<0 05) than controls (range9 16-53-23 mm2).

Physiological experiments. Chronic ligation of theductus in fetal lambs has been reported compatiblewith live birth37 although this is not the case withfetal dogs38 which survived in utero only untilshortly before parturition. When the ductus was tiedin 4 fetal lambs, the average pressure differencebetween pulmonary trunk and aorta under maternalanaesthesia rose from 1 -2 to 20-6 mmHg (0-16 to2 * 7 kPa.)37 These lambs survived for 9-36 days, theyhad 34% more ventricular muscle than littermates,and changes in the media of small muscularpulmonary arteries were found.

Administration of acetylsalicylic acid 55-90

102 Joan C Mott

mg/kg estimated body weight to catheterised fetallambs in utero increased the pressure drop frompulmonary artery to aorta from 1 * 5 to 11 *2 mmHg(0 20 to 1-49 kPa), and decreased the fraction ofright ventricular output traversing the ductus from88*5 to 69*7y%. The absolute decrease averaged15 * 9 %.' In two lambs with pulmonary hypertension,IV infusion of PGE1 0 1 mg/kg per minute restorednormal levels of pulmonary arterial pressure. Whenthe infusion was stopped pulmonary pressure roseagain.1 Substantial changes in the distribution ofcardiac output (measured with radioactive micro-spheres) were found during administration ofacetylsalicylic acid. For example, that to the liver wasdecreased by 40% and that to the lungs increased by196%.' Though blood-gas tensions were unaffected,if such larg, changes in organ flow had persisted withcontinuous treatment they might have had importantconsequences. Indomethacin (0 *005 mg/kg) not onlycaused profound ductus constriction in fetal lambsbut also reduced circulating levels of PGE.39 Infusionof PGE reversed the constriction but did not raisecirculating levels. Slow infusion of indomethacin(1 mg/kg maternal bodyweight) into pregnant ewesraised the pressure gradient across the fetal ductusfrom 2-7 ± 2 0 to 14-1 ± 5 9 mmHg during thenext 1-6 hours. The increase began within 30minutes, was not attributable to a fall of systemicpressure, but was reduced by infusion of PGE1 intwo lambs. Nine lambs which had not receivedPGE1 were found dead the next day.40

Thibeault et al.41 found that in fetal lambs treatedwith hydrocortisone for several days, the ductus wasfunctionally closed or constricted compared withcontrols. Corticosteroids can inhibit prostaglandinrelease, so that this observation would be consistentwith a role for PGE in the maintenance of ductuspatency.

Circulating prostaglandins. PGE levels are high(10-9 mol/l) in plasma obtained from catheterisedlambs in utero especially towards term.42 Levels fallin the first few days of life, as in human infantswhere they appear unaffected by gestational age atbirth.43

Discussion

Almost all in vitro work has used an anoxic (<14mmHg) relaxed ductus for the investigation ofcontractile agents, and one maximally constricted inan Oa environment of several hundred mmHg fordetection of relaxant action. Whether the apparentdifference of sensitivity to 02 of guinea-pig and lambisolated ductus is solely due to the larger size of, andgreater dependence on, vasa vasorum of the latter is

perhaps unlikely, since the even larger ductus of thecalf is contracted at a Po2 of 100 mmHg.

Clarification of such discrepancies may wellrequire more detailed anatomical knowledge anddiscrimination between the synthetic abilities of thevarious tissues comprising the ductus in variousspecies. However no doubt exists as to the gradedresponse of the ductus to increasing 02 tension bothin vitro6 9 and in vivo.2 44 The stimulation of increased0° tension remains unchallenged as a trigger forpost- (and pre-) natal constriction of the ductus.Although synthesis of prostaglandins requiresoxygen,45 it appears possible at the rather low 02tensions in the fetus.

Prostaglandins in plasma may participate in thecontrol of vascular tone in vivo in addition to thatproduced by endogenous synthesis within the vesselwall itself. In general, the behaviour in vitro ofisolated ductus tissue to exogenous prostaglandinsand to prostaglandin synthetase inhibitors accountsfor the observations in vivo. The undoubted presenceof autonomic innervation of the ductus46 has notbeen shown to play an important part in itsbehaviour.The possible roles of bradykinin29 and angiotensin

1128 in PGE synthesis may deserve closer examina-tion. However, if in the ductus, as appears both PGEand PGF2a can be produced, administration of aprostaglandin synthetase inhibitor might predomi-nantly promote constriction by inhibition of PGEproduction, or dilatation by inhibition of PGF2,production. In addition, sheep blood contains anenzyme, with a pH optimum about 7, capableof reducing PGEa to PGF2a.47 Thus the finaloutcome may not be easily predictable in the wholeorganism.The part played by prostaglandins in relation to

the contractile action of 02 on the ductus, which isbelieved to be mediated by an increased rate ofsynthesis of high-energy phosphate compounds,>8is not understood. Prostaglandins are lipid-solubleand their involvement in control of intracellularCa++ at the level of the mitochondrial membranein other systems suggests they could be responsible,48though further sites of action such as the sarcomeremembrane must remain possibilities.

Whole animal.

Fetus. If the ductus is totally (by ligation), or partly(by administration of prostaglandin synthetaseinhibitors), prevented from performing its normalfetal function there is substantial evidence thatpathological consequences can follow.3435 37-38 Anyincrease of pulmonary arterial pressure could bemitigated by increased flow through the foramen

Patent ductus arteriosus: experimental aspects 103

ovale.38 However, while brief administration ofprostaglandin synthetase inhibitors may have only ashort-lived effect on PGE synthesis, in ductus musclethe consequences of prolonged inhibition of prosta-glandin synthesis there and elsewhere cannot bediscounted. Nor can it be assumed, in the case ofcompounds which cross the placenta, that maternalplasma concentrations of the inhibitor are anyindication of those in the fetus.33 4

Newborn A histological study of human patentducts shows that all those from subjects over age 4months have abnormal subintimal elastic tissue.50This anomaly was also present at all ages from birthto 4 months together with histologically normal ductsat various stages of closure. It is therefore possiblethat congenital abnormality precludes bothspontaneous or pharmacologically-assisted closure ofsome ducts.

Tn anatomically normal newborns, which are nothypoxaemic, failure of the ductus to close could beattributable to endogenous synthesis of PGE ratherthan PGF, but there is no direct evidence on thispoint. Doubt has been expressed about whetherconcentrations of PGF2a sufficiently high to causeductus constriction are likely to occur naturally. Asfar as plasma levels are concerned, both PGE andPGF were raised in 6 of 7 human infants with patentductus arteriosus (PDA).51 In two cases a relativeincrease of plasma PGE with respect to PGFcoincided with cardiac failure due to a large left-to-right shunt through a patent ductus.52

Administration of prostaglandin synthetaseinhibitors to promote closure of the ductus cannotavoid effects on other organs. A well documentedexample is the inhibition of renal blood flow andurine flow in newborn lambs.53 Elimination ofindomethacin by premature infants54 is as slow asthat in the fetus.33

In infants in whom the pulmonary blood flow islargely dependent on a PDA and presumably onadequate endogenous synthesis of PGE, 02 levels,though low, may well exceed fetal levels and perhapshinder any dilator effect ofPGE whether endogenousor exogenous.The histology of the ductus in cases of ductus-

dependent cardiac anomaly, treated with PGE,suggested structural weakening due to intimallacerations, interruption of internal elastic laminae,and medial oedema.55 Plasma levels of PGE wereunaffected by ductus ligation.5'

In the newborn as in the adult 56-57prostaglandinsare metabolised in the lungs. The much largerpulmonary blood flow of the newborn can beexpected to dispose of any surplus prostaglandinsmore readily than the fetus.

Conclusions.

(1) The Po2 of fetal blood is insufficient to constrictthe ductus arteriosus in the face of its endogenoussynthesis of prostaglandin E.

(2) Closure of the ductus at birth is initiated by theincreased Po2 of circulating blood as pulmonaryventilation is established. The participation ofPGF2a is improbable as relatively large con-centrations are required to constrict ductusmuscle and in some experiments a dilator actionwas observed.

(3) Administration of PGE has been shown to dilatethe partly constricted postnatal ductus in vivo.

(4) Some inhibitors of prostaglandin synthetasecross the placenta; these may attain a concentra-tion in fetal plasma exceeding that in the mother,and have been shown experimentally to promoteconstriction or closure of the ductus in utero. Thehaemodynamic consequences of such closureinclude changes in the pulmonary vessels whichmay preclude the normal development ofvascularconductance after birth.

(5) If the ductus remains open postnatally atreasonable levels of oxygenation, prostaglandinsynthetase inhibitors may assist closure as itseems unlikely that any constrictor effects ofPGFsa would dominate the situation. However,the multitude of other prostaglandin synthetasesystems in the [body would also be more or lessinhibited. The possibility of adverse conse-quences of prolonged treatment requires detailedanalysis.

References

1 Heymann M A, Rudolph A M. Effects of acetylsalicylicacid on the ductus arteriosus and circulation in fetal lambsin utero. Circ Res 1976; 38: 418-22.

2 Born G V R, Dawes G S, Mott J C, Rennick B R. Theconstriction of the ductus arteriosus caused by oxygenand by asphyxia in newborn lambs. JPhysiol (Lond) 1956;132: 304-42.

3 Barcroft J, Kennedy J A, Mason M F. The relation of thevagus nerve to the ductus arteriosus in the guinea-pig(abstract). JPhysiol (Lond) 1938; 92: 1-2P.

4 Kennedy J A, Clark S L. Observations on the physio-logical reactions of the ductus arteriosus. Am J Physiol1942; 136: 140-7.

5 Koval6ik V. The response ofthe isolated ductus arteriosusto oxygen and anoxia. JPhysiol (Lond) 1963; 169: 185-97.

6 Fay F S. Guinea-pig ductus arteriosus. I. Cellular andmetabolic basis for oxygen sensitivity. Am JPhysiol 1971;221: 470-9.

7 Fay F S, Jobsis F F. Guinea-pig ductus arteriosus.Ifl. Light absorption changes during response to 02.Am JPhysiol 1972; 223: 588-95.

8 Fay F S, Nair P, Whalen W S. Mechanism of oxygeninduced contraction of ductus arteriosus. Adv Exp MedBiol 1977; 78: 123-34.

104 Joan C Mott

9 Noel S, Cassin S. Maturation of contractile response ofductus arteriosus to oxygen and drugs. Am JPhysiol 1976;231: 240-3.

10 Clyman R I, Mauray F, Wong L, Heyman M A, RudolphA M. The development and response of the ductusarteriosus to oxygen. Biol Neonate 1978; 34: 177-81.

I Olley P M, Coceani F. The in vitro response of the lambductus arteriosus to prostaglandins (abstract). Pediatr Res1972; 6: 341.

12 Coceani F, Olley P M. The response of the ductusarteriosus to prostaglandins. Can J Physiol Pharmacol1973; 51: 220-5.

13 Elliott R B, Starling M B. The effect of prostaglandinF2a in the closure of the ductus arteriosus. Prostaglandins1972; 2: 399-403.

14 Starling M B, Elliott R B. The effects of prostaglandins,prostaglandin inhibitors, and oxygen on the closure ofthe ductus arteriosus, pulmonary arteries, and umbilicalvessels in vitro. Prostaglandins 1974; 8: 187-203.

15 Coceani F, Olley P M, Bodach E. Lamb ductus arteriosus:effect of prostaglandin synthesis inhibitors on the muscletone and the response to prostaglandin E2. Prostaglandins1975; 9: 299-308.

16 Olley P M, White E P, Bodach E, Heaton J, Coceani F.The contractile response of the developing lamb ductusarteriosus to ibuprofen (abstract). Circulation 1976;53 and 54: Supplement II, 168.

17 Clyman R I, Heymann M A, Rudolph A M. Ductusarteriosus responses to prostaglandin E1 at high andlow oxygen concentrations. Prostaglandins 1977; 13:219-23.

18 Clyman R I, Mauray F, Heymann M A, Rudolph A M.Ductus arteriosus: developmental response to oxygen andindomethacin. Prostaglandins 1978; 15: 993-9.

'9 Clyman R r, Rudolph A M, Heymann M A. Ductusarteriosus: developmental response to endogenousprostaglandins (PG) and indomethacin (abstract).Pediatr Res 1979; 13: 492.

20 Coceani F, Olley P M, Bodach E. Prostaglandins: apossible regulator of muscle tone in the ductus arteriosus.In: Samuelsson B, Paoletti R, eds. Advances in prosta-glandin and thromboxane research, vol. 1. New York:Raven Press, 1976; 417-24.

21 Clyman R I, Wong L, Heymann M A, Rudolph A M.Responsiveness of the lamb ductus arteriosus to pros-taglandins and their metabolites. Prostaglandins 1978;15: 325-31.

2 2 Kriska M, Koval6ik V. The role of prostaglandins in themechanism of closure of ductus arteriosus. Acta Biol MedGer 1976; 35: 1175-6.

23 Coceani F, Olley P M, Bodach E, White E P. Significanceof the prostaglandin system to the control of muscle toneof the ductus arteriosus. In: Coceani F, Olley P M, eds.Advances inprostaglandin and thromboxane research, vol.4,Prostaglandins and perinatal medicine. New York: RavenPress, 1978: 325-33.

24 Smiesko V, Kriska M, Koval6ik V. Bayliss myogenicresponse in the isolated ductus arteriosus of guinea-pigand rabbit fetuses. Experientia 1978; 34: 745-6.

25 Pace-Asciak C R, Rangaraj G. The 6 keto prostaglandinF, x pathway in the lamb ductus arteriosus. BiochimBiophys Acta 1977; 486: 583-5.

26 Clyman R I, Mauray F, Koerpes M A, Wiemer F,Heymann M A, Rudolph A M. Formation of prostacyclin(PGI2) by ductus arteriosus of fetal lambs at differentstages of gestation. Prostaglandins 1978; 16: 633-42.

2 7 Terragno N A, Terragno A, McGiff J C, Rodriguez D J.Synthesis of prostaglandins by the ductus arteriosus ofthe bovine fetus. Prostaglandins 1977; 14: 721-7.

28 Alexander R W, Gimbrone M A, Jr. Stimulation of

prostaglandin E synthesis in cultured human umbilicalvein smooth muscle cells. Proc Natl Acad Sci USA 1976;73: 1617-20.

29 Terragno D A, Crawshaw K, Terragno N A, McGiff J C.Prostaglandin synthesis by bovine mesenteric arteries andveins. Circ Res 1975; 36 and 37: Supplement 1, 76-80.

30 Van Dorp D A. Aspects of the biosynthesis of prosta-glandins. Prog Biochem Pharmacol 1967; 3: 71-82.

31 Sharpe G L, Larsson K S. Studies on the closure of theductus arteriosus. X. In vivo effect of prostaglandins.Prostaglandins 1975; 9: 703-19.

32 Starling M B, Neutze J M, Elliott R L, Elliott R B. Studieson the effects of prostaglandins El, E2, A1, and A2 on theductus arteriosus of swine in vivo using cineangiography.Prostaglandins 1977; 12: 355-67.

3 Traeger V A, Noschel H, Zaumseil J. Zur Phar-macokinatik von Indomethazin bei Schwanzeren,Kreissenden, und deren Neugeborenen. ZentralblGynaekol 1973; 95: 635-41.

3 4Sharpe G L, Thalme B, Larsson K S. Studies on closure ofthe ductus arteriosus. XI. Ductal closure in utero by aprostaglandin synthetase inhibitor. Prostaglandins 1974;8: 363-8.

3 5Sharpe G L, Larsson K S, Thalme B. Studies on closure ofthe ductus arteriosus. XII. In utero effect of indotnethacinand sodium salicylate in rats and rabbits. Prostaglandins1975; 9: 585-96.

36 Olley P M, Bodach E, Heaton J, Coceani F. Furtherevidence implicating E-type prostaglandins in the patencyof the lamb ductus arteriosus. Eur J Pharmacol 1975; 34:247-50.

3 Ruiz U, Piasecki G J, Balogh K, Polansky B J, JacksonB-T. An experimental model for fetal pulmonary hyper-tension. Am JSurg 1972; 123: 468-71.

38 Haller J A, Jr, Morgan W W, Jr, Rodgers B M, GengosD G, Margulies S I. Chronic hemodynamic effects ofoccluding the fetal ductus arteriosus. J Thorac CardiovascSurg 1967; 54: 770-6.

39 Friedman W F, Printz M P, Kirkpatrick S E. Prosta-glandins and the fetal ductus arteriosus (abstract).Pediatr Res 1977; 11: 394.

40 Levin D L, Mills L J, Parkey M, Garriot J, Campbell W.Constriction ofthe fetal ductus arteriosus after administra-tion of indomethacin to the pregnant ewe. J Pediatr 1979;94: 647-50.

41 Thibeault D W, Emmanouilides G C, Dodge M E.Pulmonary and circulatory function in preterm lambstreated with hydrocortisone in utero. Biol Neonate 1978;34: 238-47.

42 Challis J R G, Dilley S R, Robinson J S, Thorburn G D.Prostaglandins in the circulation of the fetal lamb.Prostaglandins 1976; 11: 1041-52.

4 Mitchell M D, Lucas A, Etches P C, Brunt J D, andTurnbull A C. Plasma prostaglandin levels during earlyneonatal life following term and preterm delivery.Prostaglandins 1978; 16: 319-26.

4 4Assali N S, Morris J A, Smith R W, MansonW A. Studieson ductus arteriosus circulation. Circ Res 1963; 13: 478-89.

4 Eckenfels A, Vane J R. Prostaglandins, oxygen tension,and smooth muscle tone. BrJPharmacol 1972; 45: 451-62.

46 Kovalcik V, Kritka M, Dolezel S. The problem ofadrenergic innervation of the ductus arteriosus of theguinea-pig foetus and its role in the mechanism ofconstriction. Physiol Bohemoslov 1969; 18: 401-11.

4 Hensby C N. Reduction of prostaglandin E2 to pros-taglandin F2a by an enzyme in sheep blood. BiochimBiophys Acta 1974; 348: 145-54.

48 Carafoli E, Cravetti F. Interactions between prostaglandinE1 and calcium at the level of the mitochondrialmembrane. Arch Biochem Biophys 1973; 154: 40-6.

Patent ductus arteriosus: experimental aspects 105

4 9 Garrettson L K, Procknal J A, Levy G. Fetal acquisitionand neonatal elimination of a large amount of salicylate.Clin Pharmacol Ther 1975; 17: 98-103.

5 0 Gittenberger-de Groot A C. Persistent ductus arteriosus;most probably a primary congenital malformation.BrHeartJ1977; 39: 610-18.

51 Lucas A, Mitchell M D. Plasma prostaglandins in pre-term neonates before and after treatment for patent ductusarteriosus. Lancet 1978; 2: 130-2.

5 2 Lucas A, Mitchell M D. Prostaglandins in patent ductusarteriosus. Lancet 1978; 2: 937-8.

53 Friedman W F, Molony D A, Kirkpatrick S E. Pros-taglandins: physiological and clinical correlations. AdvPediatr 1978; 25: 151-204.

5 4 Thalji A, Yeh T F, Raval D, Pildes R S. Pharmacokineticsof indomethacin in premature infants (abstract). PediatrRes 1979; 13: 374.

55Gittenberger-de Groot A C, Moulaert A J, Harinck E,Becker A E. Histopathology of the ductus arteriosus afterprostaglandin E1 administration in ductus dependentcardiac anomalies. Br HeartJ 1977; 40: 215-20.

56 Olley P M, Coceani F, Kent G. Inactivation of pros-taglandin E1 by lungs of the foetal lamb. Experientia1974; 30: 58-9.

5 Pace-Asciak C R. Prostaglandin biosynthesis andcatabolism in the developing fetal sheep lung.Prostaglandins 1977; 13: 649-60.

JOAN C MorrThe Nuffield Institute for Medical Research,

Headley Way,Headington,

Oxford OX3 9DS