pulmonary and effect - heart · two to four bronchopulmonary segments. post-operatively, injection...

Br HeartJ 1981; 45: 133-41

Pulmonary atresia with ventricular septal defectand major aortopulmonary collateral arteriesEffect of systemic pulmonary anastomosisSHEILA G HAWORTH,* PHILIP G REES, JAMES F N TAYLOR,FERGUS J MACARTNEY, MARC DE LEVAL, JAROSLAV STARK

From the Thoracic Unit, The Hospital for Sick Children, Great Ormond Street, London

SUMMARY In nine patients with pulmonary atresia, ventricular septal defect, and major aortopulmonarycollateral arteries, the effect of a systemic-pulmonary anastomosis on the lung was studied by com-

parison of pre- and postoperative cine angiograms. Selective injections into the collateral arteries were

performed in all patients and the source ofblood supply to each bronchopulmonary segment was sought.Central pulmonary arteries were present in six children and absent in three, as confirmed at thoraco-tomy. After insertion of a shunt, central pulmonary arteries increased in size, but the intrapulmonaryvessels with which they connected remained abnormally small and were frequently stenosed. In addition,in each case the central pulmonary arteries increased flow to only five to 11 segments of lung. In theabsence of central pulmonary arteries, shunting to a hilar or a lobar pulmonary artery increased perfusionto five to nine segments of lung per case. In all these three cases the pulmonary arteries at lobar levelshowed aneurysmal dilatation proximal to a severe stenosis. Sixty-eight per cent of collateral arteries werestenosed. The findings suggest that in this anomaly, when the central pulmonary arteries are hypoplastic,the intrapulmonary branches are also hypoplastic, emphasising the need for early surgical interventionto increase blood flow while the lung still has growth potential. Further, one collateral artery may connectwith at least as many bronchopulmonary segments as does a central pulmonary artery, and the peripheralintrapulmonary arteries with which it connects may appear at least as normal angiographically as dovessels connected to central pulmonary arteries. These observations suggest that segments of lungconnected to collaterals alone should, and sometimes can, be connected to central pulmonary arteries,the aim being to produce a unifocal blood supply as a prelude to total correction.

In pulmonary atresia with ventricular septal defectand major aortopulmonary collateral arteries,recent anatomical studies have shown that theproportion of bronchopulmonary segments con-nected to central pulmonary arteries is variable andcan be a minority of segments.' Nevertheless, asystemic-pulmonary anastomosis is generally per-formed in the hope that an increase in pulmonaryblood flow will promote arterial growth and improvethe patient's clinical condition. Even if no centralpulmonary artery is present, an anastomosis may bemade to an intrapulmonary vessel.The pre- and postoperative cine angiograms of

nine patients with successful systemic-pulmonaryanastomoses were reviewed in order to assess the*Partly supported by the British Heart Foundation.Received for publication 16 July 1980

proportion of lung which had received an increasein blood flow, the appearance of the intrapulmonaryvessels, and their capacity for growth, in relationto the clinical response.

Clinical features

The nine patients described had a modified Blalock-Taussig shunt performed using polytetrafluoroethy-lene (Gore-Tex), rather than a Blalock-Taussigshunt using a subclavian artery, as is our usualpractice. To assess the results achieved with aGore-Tex shunt, pre- and postoperative cine angio-cardiograms were performed. These angiocardio-grams were used in the present study.Nine patients with pulmonary atresia, ventricular

septal defect, and major aortopulmonary collateral,33

copyright. on D

ecember 15, 2020 by guest. P

rotected byhttp://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.45.2.133 on 1 February 1981. D

ownloaded from

http://heart.bmj.com/

134

arteries had a total of 13 shunts (Table).'Fourpatients had a second shunt to the contralateral lung,the indication being shunt occlusion in one case(case 1) and progressive cyanosis despite a function-ing shunt in the other three. A modified Blalock-

Table Clinical data

Case Age at operation No. collaterals seen Centralno. preoperatively pulmonary

Right lung Left lung arteries

1 1 mth 3 2 +1 y 8 mth

2 8 mth 2 13 1 y 4mth 2 1 +4 1 y 10 mth 3 1

2 y 10 mth5 2y 3 1 -

7y6 4 y 4 1 Missed

9y7 5y 3 1 +8 6y 1 1 +9 15y 3 1 +

Abbreviations: +, demonstrated by angiography; -, not present atthoracotomy.

Haworth, Rees, Taylor, Macartney, de Leval, Stark

Fig. 2 Case 8. Selective injection into a collateralartery with a long stenosed segment.

Fig. 1 Case 2. Selective injection into a collateralartery perfusing the lingula and three segments of theleft lower lobe shows dilated and tortuous intrapulmonaryvessels in which the pressure was equal to the systemicarterial pressure.

Taussig shunt between a subclavian and pulmonaryartery was performed in 10 instances.2 3 A centralshunt from ascending aorta to the main pulmonaryartery was performed on three occasions. Gore-Texwas used for 11 shunts and woven Dacron for two.The aim of surgery was to promote pulmonary

arterial growth and development, rather than toimprove arterial oxygen saturation. After operationthere was a distinct improvement in exercisetolerance and well-being. There was, however, nosignificant change in either haemoglobin concen-tration or systemic arterial oxygen tension in thegroup.

Investigations

All patients had at least one cardiac catheterisationperformed before and after the only, or first, shuntprocedure (Table). The time interval betweensurgery and investigation varied between sevenmonths and two years (mean= 14-7 months).Biplane cine angiography was performed in allpatients, save one in whom biplane Elema roll filmswere used at the preoperative study. At the pre-operative study, collateral arteries were injectedselectively in all save two patients. A total of 18vessels was injected and the pressure in the col-lateral artery was measured in 11. Postoperatively,the shunt was injected in all studies except one.

-5wP-mb.

....

..

..... .... .... .. ..........

... .... .....

copyright. on D



j.com/

Br H


ownloaded from


Systemic pulmonary anastomosis in pulmonary atresia with major collaterals

On reviewing the cine angiograms of each casebefore and after operation, the origin, course, andbranches of the central pulmonary arteries and thecollateral arteries were mapped on a diagram of thebronchial branching pattern.4 The size and shapeof each vessel was noted, particular attention beinggiven to the presence of stenoses.An attempt was made to measure, before and after

operation, the size of the same vessel in all cases,that vessel being either an extrapulmonary, lobar, orsegmental pulmonary artery. This was abandonedbecause of the considerable variation in arterialanatomy caused by the presence of stenoses andlocalised arterial dilatations. Such abnormalitieswere generally unrelated to the site of the aorto-pulmonary anastomosis.

Results

HAEMODYNAMIC FINDINGSFourteen collateral arteries were catheterised at the

pre- and postoperative cardiac catheterisationstudies. The pressure within the collateral arteryexceeded 45/30 mmHg in all, and was equal to thesystemic arterial pressure in six. Angiocardio-graphy, however, showed that the pressure hadbeen measured proximal to a stenosis in all savethree instances. The exceptions included two col-ateral arteries in case 2, where a high collateralarterial pressure was associated with angiographicfeatures consistent with a diagnosis of pulmonaryvascular obstructive disease (Fig. 1). Central pul-monary arterial pressure was measured post-operatively in only two cases. It was equal to thesystemic arterial pressure in case 1, proximal to asevere stenosis, and measured 50/30 mmHg incase 9 without a distal stenosis.

ANGIOCARDIOGRAPHY(1) Collateral arteriesIn all nine cases a total of 36 collateral arteries en-tered the lung, 34 being shown preoperatively

Fig. 3 Case 5. Selectiveinjection into a collateral arterywhich divides to connect with foursegmental arteries of the rightlung and supplies at least sixsegmental arteries of the leftlung. The segmental arteriesappear abnormally small.Stenoses (--).

135

copyright. on D



j.com/

Br H


ownloaded from



(Table). Some collaterals branched near theirorigin from the aorta.Stenoses: these were common. Thirteen collateralshad a narrowed segment between the aorta and lung(Fig. 2 and 3). A further eight collaterals had a morelocalised constriction where the collateral anasto-mosed with an intrapulmonary artery (Fig. 4).Three collaterals had a narrow origin from the aorta.'Connections: preoperatively, collateral arterial in-jections showed central pulmonary arteries infive patients. A collateral anastomosed with theleft pulmonary artery in three cases, and the rightpulmonary artery in one. The source of supply

Fig. 4 Case 7. Selective injection into a collateralsupplying the right middle lobe and four segments of thelower lobe, showing stenoses (-*) and segmental arterialdilatation.

could not be identified by aortography alone in theremaining case (case 1). Even in the presence ofcentral pulmonary arteries, however, the majority ofcollateral arteries anastomosed end-to-end withsegmental intrapulmonary arteries, providing theonly source of blood supply to those segments, andhaving no connection either with the centralpulmonary arteries or their branches within thelung (Fig. 5a and b). In the three patients withoutcentral pulmonary arteries, the maximum numberof bronchopulmonary segments connected to onecollateral artery varied between three and eightsegments per lung.

(2) Central pulmonary arteriesPreoperatively central pulmonary arteries wereshown in five patients (Table). Subsequent thoraco-tomy disclosed central pulmonary arteries in oneadditional case, in whom the vessels had not beenshown angiographically for two reasons. Firstly,the collateral perfusing the right pulmonary arterywas severely stenosed distal to the point of injection(Fig. 6a). Secondly, retrograde flow through theright pulmonary artery was prevented by a stenosisat the origin of that vessel (Fig. 6b). A systemic-pulmonary anastomosis was performed to a centralvessel in five patients, and, m'stakenly, to an upperlobe artery alone in one.Appearance: a pre-and postoperative comparison ofthe central pulmonary arteries was possible in fourcases. All four showed a considerable increase invessel size, but in two cases the vessels still ap-peared abnormally small. Three stenoses wereshown, two of them preoperatively. In addition, twostenoses occurred around the systemic-pulmonaryanastomosis, which in cases 1 and 6 had been in-serted into an extremely small central pulmonaryartery.Connections: preoperatively, in each lung the centralpulmonary artery was seen to connect with onlytwo to four bronchopulmonary segments. Post-operatively, injection of contrast medium into thesurgical shunt rather than a collateral artery im-proved visualisation of the intrapulmonary arteries,and a further seven bronchopulmonary segmentswere seen to connect with central pulmonaryarteries. In each case, however, only five to 11segments of lung received an increase in blood flow.In case 1, a subclavian to pulmonary anastomosisinto an extremely small right pulmonary arteryoccluded the origin of two branches to the rightupper lobe.

All cases contained at least one lung in which thesource of blood supply, either central pulmonaryarterial or collateral arterial, was not demonstratedin two to five segments.

136

copyright. on D



j.com/

Br H


ownloaded from



(3) Intrapulmonary arteries connected to centralpulmonary arteriesPreoperatively, the lobar and segmental pulmonaryarteries were abnormally small and sometimes had a"spikey" appearance. At the periphery, filling waspoor. Only one stenosis was shown preoperatively,in the left lower lobe artery of case 9. Improvedpostoperative visualisation of the intrapulmonaryarteries showed additional abnormalities. In all,seven stenoses were present at either lobar orproximal segmental artery level (Fig. 5b and 6b).Aneurysmal dilatation occurred near the 6riginof the lower lobe arteries in both lungs in case 6(Fig. 6b), and dilatation of a segmental artery alsooccurred distal to a severe stenosis in case 1. It wasdifficult to assess change in size of the intrapul-monary arteries after a shunt operation. Arterialopacification was improved by injection throughthe shunt, giving what may well have been, in somecases, a spurious impression of increase in vesselsize. Postoperatively, it was the frequent demonstra-tion of stenoses in both central and intrapulmonaryarteries which was the most striking finding and onewhich made it difficult to evaluate the effect of ashunt on the vessels beyond.

(4) Intrapulmonary arteries connected to collateralarteriesPreoperatively, the intrapulmonary arteries whichwere connected to collateral vessels frequentlyappeared abnormal. Eight such intrapulmonaryarteries showed a localised dilatation at lobar orsegmental level, just distal to a narrowed segmentin the collateral artery. Beyond the site of anasto-mosis with a collateral artery, no localised con-strictions of the intrapulmonary arteries were seen.In general, the appearance of the intrapulmonaryvessels was related to the size of the collateral arteryand the presence or absence of stenoses. Pro-nounced hypoperfusion was characterised bystraight vessels with reduced branching and smallvessel haze. Pronounced hyperfusion was charac-terised by tortuous vessels with side branches andexcessive small vessel haze. In between these twoextremes came essentially normal appearances.There were also a few segments in case 2 in whichthe appearances were consistent with severe pul-monary vascular obstructive disease in that the mainvessels were dilated and tortuous, but peripheralbranching and small vessel haze were reduced.

In three patients without central pulmonary

a bFig. 5 Case 7. (a) Preoperative aortogram. (b) Postoperative angiogram, injection into aortopulmonaryanastomosis. Stenoses (->). Comparison of (a) and (b) shows a greater proportion of lung connected to collateralthan to central pulmonary arteries.

137

copyright. on D



j.com/

Br H


ownloaded from



arteries, the shunt was inserted into a hilar or lobarpulmonary vessel which connected with eitherfive, seven, or nine bronchopulmonary segments.In all cases, aneurysmal dilatation of the pulmonaryarteries at lobar level occurred proximal to a stenosisbetween these and the segmental vessels. The sizeof the distal intrapulmonary arteries varied. Inboth cases 2 and 5 a shunt increased flow to fourabnormally small segmental arteries. In case 4,however, a shunt to an intrapulmonary artery of theright lung increased perfusion to seven segmentalarteries of normal size.

In patients with central pulmonary arteries, thosesegments of lung connected to a collateral arteryalone often contained larger arteries with morelateral branching and more small vessel haze thandid those segments connected to the central pul-monary artery, despite the fact that a shunt hadbeen performed (Fig. 5a and b).

Ligation of collateral arteries in patients withcentral pulmonary arteries occluded the flow to twosegments of the right lower lobe in case 1 and fivesegments of the right upper and lower lobes incase 3.

Discussion

In order to develop a rational surgical approach tothe management of pulmonary atresia, it is essentialto ask what is meant by surgical correction. Repairof the cardiac defect alone is in many cases in-adequate, for reasons which will be elaborated later.We shall argue that the pulmonary blood supplyalso needs correction, but this poses further ques-tions. What precisely needs correcting? What isnormal about pulmonary blood supply in this.condition, and what is not? And if abnormalitiesexist, to what extent are they (a) reversible and (b)correctable?A clear answer to the questions has been madeX

more difficult by lack of clarity in three key areas.Firstly, sources of pulmonary blood supply presentat birth have not always been distinguished fromacquired collateral circulation, particularly instudies based on older children. Secondly, there hasbeen uncertainty as to the nature of the largesystemic arteries perfusing the lungs in this condi-tion, which has in turn led to confusion about thenature of the segments of lung they supply. For if

a b

Fig. 6 Case 6. (a) Preoperative injection into stenosed collateral artery (single arrow) perfusing four segments of theright lung. The position of the stenosed origin of the right pulmonary artery is also indicated (double arrow) and isshown in Fig. 6b. (b) Postoperative injection into aortopulmonary anastomosis, showing central pulmonary arteries,stenoses (-*), and aneurysmal dilatation at origin of left upper and lower lobe arteries.

138

copyright. on D



j.com/

Br H


ownloaded from



these systemic arteries are termed bronchialarteries, the presumption is that the intrapulmonaryarteries are also bronchial arteries, and that nonormal intrapulmonary arteries exist in thesesegments. If, on the other hand, the large systemicarteries are termed aortopulmonary collateralarteries, the implication is that they are abnormalconnections between essentially normal systemicand normal intrapulmonary arteries. And if normalintrapulmonary arteries exist, complete surgicalcorrection would involve restoration of continuitybetween the central (intrapericardial) and allperipheral (intrapulmonary) pulmonary arteries.Thirdly, the words "pulmonary artery" or even"the pulmonary artery" have sometimes been usedunqualified, leaving uncertainty as to whetherreference is being made to central or peripheralvessels.

In many of the patients in the present study somedegree of acquired collateral circulation developedbetween the first and second angiocardiographicstudy on the side of the thoractomy. This appearedas fine, tortuous arteries arising principally fromintercostal arteries and having no direct connectionwith hilar or intrapulmonary arteries, in contrast tocongenital collateral circulation. Acquired collateralcirculation presumably increases effective pul-monary blood flow, since the anastomosis is atprecapillary level,5 but is otherwise disadvantageous,since it will increase pulmonary venous return dur-ing cardiopulmonary bypass in an uncontrollablefashion. Because in our cases acquired collateralblood supply formed an unquantifiable but probablysmall proportion of total pulmonary blood flow, itwill not be discussed further.There is now considerable evidence from histo-

logical,1 6 radiographic,7 8 physiological,8 and em-bryological9 10 studies that the large systemic arterieswhich are present at birth and provide the pul-monary blood supply are not bronchial arteries,which is why we have termed them major aorto-pulmonary collateral arteries. The same argumentslead to the conclusion that the intrapulmonaryarteries in this condition are in essence normal,' 8developing and having the same course, relations,and structure as centrally connected vessels. Thisstudy, however, while confirming the generalvalidity of this approach, indicates, as did a previousnecropsy study,' that to be in essence normal doesnot necessarily mean to be in fact normal. Apartfrom the previously well-documented stenoses inmajor aortopulmonary collateral arteries,7 811 12obstructions were also seen in both central andperipheral pulmonary arteries in the hilar region,much as described by McGoon and colleagues."The severity of these stenoses was not apparently

affected by shunt operations, and complete surgicalrepair would involve relief of such obstructions. Inaddition to stenoses, localised dilatations of intra-pulmonary arteries at lobar or segmental level werenot infrequent, either immediately distal to a majoraortopulmonary collateral artery stenosis or proxi-mal to a lobar pulmonary artery stenosis. Thesedilatations are probably of little haemodynamicsignificance, but make it difficult to quantify theeffect of shunt operations on the growth of thecentral pulmonary arteries and their continuationsin the lung. However, in five patients there was adistinct increase in the size of the central pulmonaryarteries after a shunt.There was also an increase in perfusion at the

periphery, but it is uncertain to what extentshunting the central pulmonary artery had reversedthe effect of hypoperfusion on lung development.In young children, improving blood flow to under-perfused areas of lung should improve the growthof intra-acinar arteries in terms of both the size andnumber of vessels."3 Given that in some patientsperipheral perfusion appeared, however, to be lessin some shunted segments than in segments per-fused by collaterals, it does seem possible thatshunting may not always restore the peripheralvasculature to normal. In the present study, thesegmental and peripheral pulmonary arteries ap-peared abnormally small, even when distended byblood and contrast medium, while in the previousmorbid anatomical study,' vessels at this level wereabnormally small in relation to the size of the ac-companying airway, irrespective of whether theywere normally connected or isolated. In bothstudies, the problem within the lung was usuallyone of hypoplasia. Microscopical quantitativemorphometric analysis' has shown a reduction insize and muscularity of the intra-acinar arteries andeccentric areas of intimal proliferation, structuralchanges compatible with long-standing reducedpulmonary blood flow.'4The major problem, however, is that not enough

intrapulmonary arteries are connected to thecentral pulmonary arteries. In contrast to thenormal situation when the central pulmonaryarteries form a single common pressure head (orfocus) for perfusion of the lungs, in pulmonaryatresia with ventricular septal defect and majoraortopulmonary collateral arteries there is nearlyalways more than one distinct pressure head so thatthe pulmonary blood supply is multifocal.There are now many reports of pulmonary

hypertension occurring immediately after surgicalcorrection of pulmonary atresia with ventricularseptal defect.15-17 This is indeed one of the mostimportant factors affecting immediate surgical

139

copyright. on D



j.com/

Br H


ownloaded from


Haworth, Rees, Taylor, Macartney, de Lavel, Stark

mortality.'7 It has been suggested for some timethat the most important underlying cause for this ismultifocal pulmonary blood supply,5 since if theentire cardiac output is forced to pass through asufficiently limited quantity of pulmonary vascularbed, pulmonary hypertension is inevitable. Thistheoretical concept has recently received strongsupport from a study in which the ratio betweenright and left ventricular systolic pressures, im-mediately after surgical correction of pulmonaryatresia with ventricular septal defect, was predictedfrom the ratio between the diameter of the proximalright and left pulmonary arteries and the aorta, thebody surface area, the presence of stenosis of theright pulmonary artery, and the presence ofarborisation anomalies in the pulmonary vascularbed.'7 This showed that the predicted RV/LVsystolic pressure ratio should be raised by 0 85 ifbilateral arborisation anomalies are present, bywhich the authors meant that one lobe or more oneach side was not normally connected to a centralpulmonary artery.

It is noteworthy that this degree of arborisationabnormality was present in all patients with pul-monary atresia and ventricular septal defectdescribed here, as well as in nearly all patients in arecent necropsy study from this institution.' Whileit is true that this is a group selected to some extentby the decision not to proceed directly to completerepair, nine other patients underwent "complete"repair during the three years spanning the operationdates for these shunted patients. All survivedoperation but only two had pulmonary atresia withventricular septal defect and major aortopulmonarycollateral arteries, and one of these died later frominfection. In our experience, therefore, the majorityof patients with complex pulmonary atresia have asevere arborisation anomaly, and by the criteria ofAlfieri and colleagues'7 are unlikely to surviveconventional corrective surgery with an acceptableRV/LV systolic pressure ratio. If the arborisationabnormality were also corrected, such patients mightsurvive repair of the cardiac defect.The results of this study indicate that shunting

does not correct the connection problem. Selectiveinjections into isolated major aortopulmonarycollateral arteries (those not connected to centralpulmonary arteries) showed no change in ap-pearance between the pre- and postoperative studies,indicating that in no case had a shunt into a centralpulmonary artery resulted in establishment ofperfusion of bronchopulmonary segments notconnected to central pulmonary arteries. There wasno evidence of an anatomical communication be-tween collaterally connected and normally con-nected intrapulmonary arteries. The possibility of

such communications has not been ruled out,because we have noticed in these and other patientsthat during selective injection into one majoraortopulmonary collateral artery it is occasionallypossible to identify "washout" in the lobar pul-monary arteries where non-opacified blood entersfrom another lobar artery supplied by a differentmajor aortopulmonary collateral artery. In theanatomical study,' however, such communicationswere sought but not found. Furthermore, if suchconnections were frequent, it would be exceptionalto find side by side an overperfused pulmonaryhypertensive segment and an underperfused seg-ment. In reality, this is common.

Given this basic compartmentalisation of pul-monary blood supply, it is clear that however muchblood can be forced through the central pulmonaryarteries, whether by systemic-pulmonary shuntsalone, as here, or by placing conduits between theright ventricle and pulmonary artery without closingthe ventricular septal defect,'8 the "connectionproblem" will not be solved.Two possible ways exist of making multifocal

blood supply unifocal. One is to ligate isolatedmajor aortopulmonary collateral arteries, in order toencourage compensatory growth of normally con-nected lung segments,'9 as the underperfusedsegments might become relatively smaller. In thesmall number of cases where this had been carriedout, the effect on the total volume of lung normallyconnected was not detectable. It is preferable toexplore the possibility of establishing direct con-tinuity between all segments of lung by creatinganastomoses between lobar pulmonary arteries inthe region of the hilum, and overcoming the stenosescommonly present at this level. For technicalreasons, this operation is carried out during theshunt procedure rather that at the time of completerepair. A division of a collateral artery with inter-ruption of the aortic end and anastomosis of thedistal end, either directly or using a Gore-Texprosthesis, to the central pulmonary artery, isanother alternative. We have performed thisoperation in four patients. This will be the subjectof a further report when the follow-up angiogramsbecome available.

References

1 Haworth SG, Macartney FJ. Growth and develop-ment of pulmonary circulation in pulmonary atresiawith ventricular septal defect and major aorto-pulmonary collateral arteries. Br Heart J 1980; 44:14-24.

2 McKay R, de Leval MR, Rees P, Taylor JFN,Marcartney FJ, Stark J. Postoperative angiographicassessment of modified Blalock-Taussig shunts

140

copyright. on D



j.com/

Br H


ownloaded from



using expanded polytetrafluoroethylene (Gore-Tex).Ann Thorac Surg 1980; in press.

3 de Leval MR, McKay R, Jones M, Stark J,Marcartney FJ. Modified Blalock-Taussig shunt.Use of the subclavian artery orifice as a flow regu-lator in prosthetic systemic to pulmonary arteryshunts. Jf Thorac Cardiovasc Surg 1980; in press.

4 Foster-Clark AF. Broncho-pulmonary anatomy.In: Perry KMA, Holmes Sellors T, eds. Chestdiseases. London: Butterworths, 1963; 1: 19, 21-22.

5 Macartney FJ. The haemodynamics of the abnormaloutflow tract. In: Dyde JA, Smith RE, eds. Surgeryof the heart. New York: Plenum Publishing Corpora-tion, 1976: 103-17.

6 Thiene G, Frescura C, Bini RM, Valente M,Gallucci V. Histology of pulmonary arterial supplyin pulmonary atresia with ventricular septal defect.Circulation 1979; 60: 1066-74.

7 Jefferson K, Rees S, Somerville J. Systemic arterialsupply to the lungs in pulmonary atresia and itsrelation to pulmonary artery development. Br HeartJ 1972; 34: 418-27.

8 Macartney FJ, Deverall PB, Scott 0. Haemody-namic characteristics of systemic arterial bloodsupply to the lungs. Br Heart J 1973; 35: 28-37.

9 Congdon ED. Transformation of the aortic-archsystem during the development of the humanembyro. Contributions to Embryology. CarnegieInstitute 1922; 14, nos. 65-71: 47-112.

10 Boyden EA. The time lag in the development ofbronchial arteries. Anat Rec 1970; 166: 611-4.

11 McGoon MD, Fulton RE, Davis GD, Ritter DG,Neill CA, White RI Jr. Systemic collateral andpulmonary artery stenosis in patients with congenitalpulmonary valve atresia and ventricular septaldefect. Circulation 1977; 56: 473-9.

12 Levin DC, Baltaxe HA, Goldberg HP, et al. The

importance of selective angiography of systemicarterial supply to the lungs in planning surgicalcorrection of pseudotruncus arteriosus. AJR 1974;121: 606-13.

13 Hislop A, Reid L. Pulmonary arterial developmentduring childhood: branching pattern and structure.Thorax 1973; 28: 129-35.

14 Best PV, Heath D. Pulmonary thrombosis incyanotic congenital heart disease without pulmonaryhypertension. Jf Pathol Bacteriol 1958; 75: 281-91.

15 Somerville J, Ross D. Long-term results of completecorrection with homograft reconstruction in pul-monary outflow tract atresia. Br Heart Jf 1972;34: 29-36.

16 Doty DB, Kouchoukos NT, Kirklin JW, Barcia A,Bargeron LM Jr. Surgery for pseudotruncusarteriosus with pulmonary blood flow originatingfrom upper descending thoracic aorta. Circulation1972; 45, suppl 1: 121-9.

17 Alfieri 0, Blackstone EH, Kirklin JW, Pacifico AD,Bargeron LM Jr. Surgical treatment of tetralogy ofFallot with pulmonary atresia. J Thorac CardiovascSurg 1978; 76: 321-35.

18 Gill CC, Moodie DS, McGoon DC. Staged surgicalmanagement of pulmonary atresia with diminutivepulmonary arteries. Jf Thorac Cardiovasc Surg 1977;73: 436-42.

19 Macartney FJ, Scott 0, Deverall PB. Haemody-namic and anatomical characteristics of pulmonaryblood supply in pulmonary atresia with ventricularseptal defect-including a case of persistent fifthaortic arch. Br HeartJ 1974; 36: 1049-60.

Requests for reprints to Dr S G Haworth, ThoracicUnit, The Hospital for Sick Children, GreatOrmond Street, London WC1N 3JH.

141

copyright. on D



j.com/

Br H


ownloaded from


pulmonary and effect - heart · two to four bronchopulmonary segments. post-operatively, injection...

Documents