pulmonary vascular and development in tetralogy a correctiontetralogy in patients who usually...

Thorax 1982;37:893-901

Pulmonary vascular and alveolar development intetralogy of Fallot: a recommendation for earlycorrectionRJ JOHNSON, SHEILA G HAWORTH

From the Department ofPaediatric Cardiology, Institute of Child Health, London

ABSTRACT Using quantitative morphometric techniques, we analysed pulmonary arterial andalveolar development in the lungs of seven children aged 1 2-12 years who died during or soonafter repair of tetralogy of Fallot. One child had a residual ventricular septal defect and survived forfive months. One other child had had a previous Waterston-Cooley anastomosis (Waterstonshunt). Postmortem lung volume in relation to body surface area was generally below normal forage, the alveoli were small, and the total alveolar number was below normal in five of the sevencases. Microscopically, airway and alveolar structure appeared normal. The preacinar arterieswere larger and the intra-acinar arteries were smaller than normal for age. The preacinar elasticpulmonary arteries appeared to contain less elastin and in both preacinar and intra-acinar musculararteries the media was thinner than normal, although muscle was normally distributed along thearterial pathway. Eccentric areas of intimal fibrosis were small and uncommon. The bronchialarteries were generally more prominent than usual both macroscopically and microscopically, butno abnormal bronchopulmonary connections were present. After corrective surgery a residualventricular septal defect and pulmonary hypertension were associated with arterial medialhypertrophy, and this change was also found in the right lung of a normotensive patient who hadhad a Waterston shunt. This group probably represents the most favourable clinical picture oftetralogy in patients who usually survive but, even so, pulmonary arterial and alveolardevelopment was abnormal. The structural findings are discussed in relation to the functionaloutcome in patients with tetralogy who have survived. Repair of the abnormality during the firsttwo to three years of life is recommended.

For most patients with tetralogy of Fallot theprognosis is good. The five-year survival of patientswho have had a Blalock-Taussig anastomosisfollowed by "corrective" surgery is 85% and manychildren undergo elective repair without havingneeded palliative surgery.' In such patients theextrapulmonary arteries are not appreciablyhypoplastic, but within the lung the state ofdevelopment of the pulmonary circulation isuncertain. At corrective surgery, however, the abilityof the lung to accommodate an increase in blood flowsatisfactorily depends on the capacity of thepulmonary circulation. In adult life the structure andfunction of the lung is probably related to the state of

Address for reprint requests: Dr SG Haworth, Department ofPaediatric Cardiology, Hospital for Sick Children, Great OrmondStreet, London WC1N 3JH

893

pulmonary vascular development at the time ofcorrective surgery. Previous studies on the structureof the pulmonary vasculature have given conffictingresults, probably because of the variation in theclinical picture and in the age of the patientsstudied.2"

In the present study pulmonary vascular andalveolar development have been analysed in the lungsof seven children who died during or soon aftercorrective surgery. Six patients had not requiredpalliative surgery and they probably represent themost favourable clinical picture of tetralogy inpatients who usually survive.

Patients and methods

THE PATIENTSSeven patients aged 1P2-12 years died after surgical

copyright. on F

ebruary 13, 2020 by guest. Protected by

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.37.12.893 on 1 Decem

ber 1982. Dow

nloaded from

http://thorax.bmj.com/

Johnson, Haworth

repair of tetralogy of Fallot (table 1). In one child(case 7) a Waterston shunt had been performed at 21/2years of age. None of the other patients had hadpalliative surgical treatment and they will thereforebe described as untreated. All were on or below the10th centile for height except for one child (case 7),who was on the 50th centile. All patients had reducedexercise tolerance and all but one (case 5) werecyanosed.

In all patients the preoperative chest radiographshowed a heart of normal size (cardiothoracic ratio0 43-0- 52), with a reduction in peripheral pulmonaryvascular markings. On pulmonary angiography theextrapulmonary, hilar, lobar, and segmental arteriesappeared to be of normal size, although theperipheral branches were smaller than normal (fig 1).

Five patients died during or within 12 hours ofsurgery, four with bleeding problems. In cases 5 and 6

Fig 1 Pulmonary angiogram (case 2) showing hilar andlobar arteries ofnormal size and smallerperipheral arteries.

there was a residual ventricular septal defect andthese children died 10 days and five months after theoperation.

PREPARATION OF LUNG TISSUE ANDMORPHOLOGICAL ANALYSISAt necropsy the pulmonary arteries were injectedwith a barium sulphate and gelatin mixture at 60°Cand a pressure of 75 6 mm Hg.9I0 The lungs werefixed by injecting the airways with formol saline at apressure of 45 cm H2O. The volume of each fixed lungwas determined by water displacement.On a postmortem arteriogram the length of the

posterobasal artery from the hilum to the pleuralsurface was measured, and the lumen diameter'wasmeasured at the hilum and at 25% intervals along thepathway. Most of the children were smaller thannormal for their age, and to compensate for this thelumen diameter at each position in each lung wasrelated to the total length of the pathway and thisratio compared with that found in controls of similarage, Student's paired t test being used. For the group,to determine the percentage change from normal ananalysis of variance was carried out by comparing thelumen diameter at each position with the normal forage, the total pathway length being used as thecovariate.To determine the number of alveoli in the lung the

volume proportion of lung parenchyma wasdetermined by a point-counting technique." Blocksof tissue were selected for histological preparation bya randomised sampling technique. Sections of tissuewere stained with haematoxylin and eosin and withMiller's elastic stain counterstained with van Gieson'sstain.

MICROSCOPIC STUDIESThe structure of the elastic pulmonary arteries wasexamined at hilar and lobar level in six cases. Thenumber of elastic laminae was determined at threepoints around the circumference of each vessel by the

Table 1 Clinical data on the seven cases oftetralogy ofFallot

Case no Age (y) Pubnonary Pa Systemic Pa Systemic SaO (%) Hb (gidl)(mm-Hg) (mm Hg)

Immediate postoperatve deaths1 1-2 12/6 100/70 66 17 72 2-6 20/9 86/46 89 14 93 3 8 16/6 105/75 60 1544 4 4 10/0 90/70 87 14 05 12 15/5 75/50 92 15-0Late postoperatve death (residual ventricular septal defect)6 8 3 65/30 - 96 12-1Previous Waterston shunt7 6 4 18/10 140/90 66 14 7

Pa-arterial pressure; Sao2-arterial oxygen saturation.

894

copyright. on F



/T


ber 1982. Dow

nloaded from


Lung growth in tetralogy of Fallot

linear intercept technique and the mean number ofelastic laminae was calculated. In each peripheralpulmonary artery examined the external diameterand medial wall thickness were measured. Percentagemedial thickness was calculated as (2 x medial wallthickness) . external diameter x 100. Arteries weregrouped into size ranges, with external diameters of0-25, 26-50, 51-75, 76-100, 101-200, 201-300, 301-400, 401-500, 501-1000, and greater than 1000 ,um.For each size range mean percentage arterialthickness and the standard error of the mean werecalculated.

Intra-acinar arteries were identified according tothe type of peripheral airway they accompanied. Theexternal diameter of arteries accompanying terminaland respiratory bronchioli, alveolar ducts, andalveolar walls was measured and the mean externaldiameter was calculated for each airway level. Theproportions of muscular, partially muscular, and non-muscular arteries at each airway level were alsodetermined.

In each lung a population count was done bymeasuring the external diameter and determining thewall structure (muscular, partially muscular, or non-muscular) of all the intra-acinar arteries in eachsection of tissue examined. For each lung the medianexternal diameter of the partially muscular arterieswas calculated. To determine how far along thepathway muscle extended the median externaldiameter of the partially muscular arteries was

compared with the mean external diameter of arteriesaccompanying respiratory bronchioli and alveolarducts in each lung.The numbers of arteries and alveoli lying within the

same microscopic field were counted. In eachspecimen a minimum of 25 fields was examined. Theratio of alveoli to arteries was used to assess arterialnumber in order to compensate for differences in the

degree of inflation. The microscopic volume pro-portion of alveolar air space was determined. Thenumber of alveoli per cubic centimetre and the totalnumber of alveoli in the lung were calculated with theformula derived by Dunnill ":

n3/2N=~~xPxV,

where N = total alveolar number, n = number ofalveoli per cm2, P = volume proportion ofparenchyma, p = volume proportion of alveolar air,V = lung volume, 16 = alveolar shape constant of1-55.

In each case the findings on the postmortemarteriogram; the results of the microscopic analysis ofarterial muscularity, size, and number; and thealveolar number were compared with publishedfindings in a child of similar age.'2 The difference inage between the four youngest patients and thecontrols was less than three months and for the olderchildren the difference was between one and twoyears. The control lung specimens had beenexamined by exactly the same methods as in thepresent study and were obtained from children whohad died accidentally.

Results

LUNG VOLUME AND NUMBERS OF ALVEOLIIn all the cases examined lung volume was smallerthan normal with respect to both age and bodysurface area (table 2). The mean lung volume was

75% of normal for age. In all cases the number ofalveoli per cubic centimetre was greater than normalwith respect to both age and body surface area. Thetotal number of alveoli in each lung examined was

normal in only two children (cases 1 and 7) andreduced in the other five cases. Thus despite an

Table 2 Lung volume and numbers ofalveoli in the cases oftetralogy ofFallot (normal values for age in parentheses)

Case No Age (y) Body surface Lung Lung Alveoli per Total No ofalveoli Alveolar:area (m2) examined volume* cm3 x 103t X 106t arterial ratio

1 1*2 038(5) Right 160(285) 1080(284) 174(143) 8-2Left 158 (280) 1100 (284) 174 (143) 8-4

2 2-6 0o42 (25) Left - - - 4-83 3-8 0-51(2-5) Right 342(380) 518(367) 178(252) 8-0

Left 287 (300) 667 (367) 192 (252) 8-54 4-4 0-56 (2 5) Right 295 (380) 623 (367) 184 (252) 5-5

Left 188 (300) 621(367) 117 (252) 6-55 12 099(2-5) Left - - - 8-66 8-3 0-82(10) Left 388(1065) 509(243) 197(290) 6-97 6-4 0-92 (50) Right 329(1185) 895 (367-243) 294 (252-290) 8-0

Left 319 (1180) 732 (367-243) 233 (252-290) 10-5(8-lot)

*Normal values from Hislop 12 in parentheses.tNormal values derived from Dunnill " in parentheses.

895

copyright. on F



/T


ber 1982. Dow

nloaded from


Johnson, Haworth

increase in the density of alveoli the reduction in lungvolume led to a reduction in the total alveolar numberin most cases.

ARTERIOGRAMSIn all seven cases the pulmonary arterial branchingpattern was normal and there were no stenoses. Bycomparison with age-matched controls, the lumendiameter of the preacinar arteries was increased alongthe entire length of the pathway in all specimensexcept the left lung in case 7, where it was normal. Inaddition, the lumen diameter related to total pathwaylength at each position between the hilum and thepleural surface was greater than that in age-matchedcontrols in all cases (fig 2), and for the group theincrease was statistically significant at the 25%position (p < 0 05) (fig 3). In the child (case 7) whohad had a Waterston shunt, along the proximal 60%of the arterial pathway the lumen diameter wasgreater in the right than in the left lung but moredistally the lumen diameter was less in the right thanin the left lung. The difference in the gradient of thelumen diameter between hilum and pleural surface inthe two lungs was significantly different (p < 0-01).The bronchial arteries filled with the injectionmedium and were larger than normal in five of theseven cases.

MICROSCOPIC FINDINGSIn all cases the airways and alveoli had a normalappearance. The small preacinar arteries appeareddilated by comparison with the accompanyingairway. These vessels also branched more frequentlyand at a more obtuse angle than is seen in the normallung (fig 4). The bronchial arteries running with thelarge airways were distended with the injectionmedium to a greater extent than is normal and theyappeared larger than normal in four cases.Bronchopulmonary anastomoses (the connectionsbetween large bronchial arteries and adjacent thin-walled arteries of the alveolar ducts and alveolarwalls 13) were more prominent and larger than in thenormal lung. On serial sectioning, bronchial arterieswere traced to the necks of the terminal andrespiratory bronchioli, where they merged with thin-walled alveolar duct and alveolar wall arteries, as isnormal. Both the preacinar and the intra-acinararteries appeared to be abnormally thin walled.

APPEARANCE OF HILAR AND LOBAR ELASTICINTRAPULMONARY ARTERIESBy definition, in the normal lung an elastic artery iscomposed of seven or more concentric elastic laminaewith muscle cells and collagen lying between the

16

112.

c~O

0*4-

02 Normal

0 25 50 75Distcace from the hilun (%)

Fig 2 Relation oflumen diameter to total length ofarterialpathway: (lumen diameter . totalpathway length) x 10,determined at intervals along the arterialpathway inspecimensfrom allseven cases (except the left lung in case 7),compared with the normal valuesfor the youngest(A- - - -A) and oldest patient (+- - - +).

A- AU-.o-

} case I

case 2} case 3

x } case 4* - case SN-19 case 6- O case 7 (right lung).

elastic fibres.'4 In five of the six cases of tetralogy inwhich elastic wall structure was examined, thearteries near the hilum contained fewer elasticlaminae than is usual at the hilum-from six to nine.The lobar arteries contained five or six elasticlaminae. In case 1, the exception, the rightpulmonary artery contained 20 elastic laminae. Thusin most cases the hilar arteries just fulfilled the criteriaof an elastic artery, and the proximal region of thelobar arteries had the structure of transitionalarteries, normally present at a more distal point alongthe arterial pathway, midway between hilum andpleural surface. Furthermore, the elastic fibres werethinner and more fragmented than normal and therewere fewer interlamellar thin elastic fibres thannormal, particularly in cases 2, 3, and 4.

In case 7, where there had been a Waterstonanastomosis, the elastic pulmonary arteries in the leftlung were thin walled, as in the other cases oftetralogy, but were thicker in the right lung (fig 5).The pulmonary artery of the right lower lobe wascomposed of only six elastic laminae; but it containeda greater amount of muscle between the laminae thandid the same vessel in other cases, and theinterlamellar distance was greater.

896

copyright. on F



/T


ber 1982. Dow

nloaded from



Pre -acinar arteriographic lumendiameter

Intra-acinar: microscopic meanexternal diameter

0 25% 50%Distance trom hilum

Fig 3 Comparison ofpreacinarand intra-acinararterial size in seven cases oftetralogy (11 lungspecimens) with the valuesfor age-matched normal controls. For thegroup the lumen diameter ofthe preacinar arteries in relation to totalpathway length was greater and the mean externaldiameter ofthe intra-acinar arteries was less than normalfor age. Vertical bars represent thestandard error.

INTIMAL CHANGEIntimal changes were neither severe nor wide-spread. None were found in the lungs of the twoyoungest patients (cases 1 and 2). In the remainingfive cases only small eccentric plaques of intimalfibrosis were present in a minority of small preacinar,terminal, and respiratory bronchiolar and alveolarduct arteries. The child who died with a residual

A~~~~~~~~~~~~~~~~~Au~~~~~~~~~~~~~~~~~ E

.+ ov, ~ ~ ~ ~ tk

Fig 4 Dilated thin-walled preacinar and small intra-acinararteries. PrA-preacinar artery; RbA-respiratorybronchiolar artery. (x 62.)

ventricular septal defect (case 6) showed the great-est intimal changes, but they were not severe.

QUANTITATIVE ASSESSMENT

Arterial muscularityThe mean percentage arterial medial thickness of theeight lung specimens analysed in cases 1-5 was

significantly less than normal (fig 6). In only two ofthese specimens was muscularity normal, and in bothinstances it was significantly reduced in thecontralateral lung. Of the two remaining patients,muscularity was significantly increased in the childwho died with a residual ventricular septal defect andpulmonary hypertension (case 6), and in the rightlung of the child who had had a Waterston shunt (case7). In this case in the left lung muscularity wassignificantly reduced as in the other cases of tetralogy.Comparison of the median external diameter of the

partially muscular arteries with the mean externaldiameter of arteries accompanying peripheralairways showed that in all except two cases muscleextended as far along the arterial pathway as isnormal in relation to age-namely, to alveolar ductlevel. Muscle extended further than normal along the

*30

+20

+10

-10

20L-UsE20)6u

0)

-c

~~~~~~6~~~6

._ u

'a aO 2.E 0. ~

75%

-20

-30

-40'

897

copyright. on F



/T


ber 1982. Dow

nloaded from


Johnson, Haworth

Fig 5 Large elastic pulmonary artery from (a) the left lung and (b) the right lung of the child (case 7) who had had aWaterston-Cooley anastomosis, showing the greater wall thickness ofthe right intrapulmonary arteries. (x 450.)

pathway into smaller, more peripheral arteries in case6 (pulmonary hypertension) and in the right lung incase 7 (Waterston shunt).

Fig 6 shows that muscle is present in smallerarteries than normal in children who had a reductionin relative medial thickness. Examination of the sizeand structure of arteries accompanying peripheralairways shows that this is because the arteries are

smaller than normal.

9.

8-

-o7

a, 6-

.w5

4.

, 3.

2-

1-1

v rI T I I

100 200 300 400 500 600 700 8O0 900 1000External dianeter (,um)

Fig 6 Mean percentage arterial medial thickness in relationto external diameter (>.m) in cases oftetralogy comparedwith normal. Cases 1-5 and left lung ofcase 7 0 0; case

6A A; case 7right lung *- *; normal 0---- -O.The vertical bars represent the standard error. + Significantdifference from normal (p < 0 05).

Intra-acinar arterial sizeUnlike the preacinar arteries seen on thearteriograms the intra-acinar arteries were smallerthan normal in five of the seven cases. For the groupof seven cases the mean external diameter of arteriesaccompanying terminal and respiratory bronchioliand alveolar ducts was significantly below the normalmean size for age (p < 0 01-4) 001) (fig 3). There wasno tendency for size to decrease further from normalwith increasing age. Intra-acinar arterial size wasnormal in cases 5 and 6. In the four cases in which acomparison between the right and left lung could bemade the arteries were of similar size in both lungs,and this includes the case (no 7) in which there hadbeen a Waterston-Cooley anastomosis.

Number ofintra-acinar arteriesOn the postmortem arteriograms an increase in thedensity of the background haze in all except that ofcase 6 suggested an increase in the number of intra-acinar arteries. This was confirmed by counting thenumber of arteries and alveoli per unit area of lung.The alveolar:arterial ratio was below normal in threecases and at the lower limit of normal in all theremaining specimens except the left lung in case 7(table 2).

SUMMARY OF PATHOLOGICAL FINDINGSLung volume related to body surface area wasgenerally below normal for age, the alveoli weresmall, and the total alveolar number was smaller thannormal in five of the seven cases. Microscopically,airway and alveolar structure appeared normal. Thepreacinar arteries were larger and the intra-acinararteries were smaller than nonnal for age. Thepreacinar elastic pulmonary arteries appeared to

898

lu

ONC-0=-iwr

nI

copyright. on F



/T


ber 1982. Dow

nloaded from



contain less elastin and in both preacinar and intra-acinar muscular arteries the media was thinner thannormal, although muscle was normally distributedalong the arterial pathway. Eccentric areas of intimalfibrosis were small and uncommon. The bronchialarteries were generally more prominent than usualboth macroscopically and microscopically, but noabnormal bronchopulmonary connections werepresent. After corrective surgery a residual ven-tricular septal defect with pulmonary hyper-tension was associated with arterial medialhypertrophy, and this change was also found in theright lung of a normotensive patient who had had aWaterston shunt.

Discussion

In patients with tetralogy of Fallot who have notundergone previous palliative surgery the findings inthe present study suggest that after an intracardiacrepair an increased volume of blood is ejected intolarge, distensible preacinar arteries which branch intoan increased number of small, thin-walled intra-acinar arteries. Although these structural abnor-malities must alter the flow profile, they would seemunlikely to increase resistance to flow. Theseobservations support the favourable clinical reportsof cardiovascular function after repair. The absenceof pulmonary hypertension in 101 patientsinvestigated by Gotsman et al5 and by Epstein et al'6attests to the capacity of the pulmonary vascular bed.

In the present series, pulmonary angiography in lifeshowed hilar, lobar, and segmental arteries of at leastnormal size and postmortem arteriography showedthese arteries to be dilated. The small preacinararteries were also dilated on microscopicexamination. Arterial dilatation was also found byWagenvoort et al,7 " but the finding is not readilyexplained. In the present study the increase in size onthe postmortem arteriogram cannot be attributed tothe injection technique since Hislop and Reid8 usedthe same technique and showed a reduction in arterialsize in a group of patients with tetralogy. Arterialdilatation might, however, be due to structuralchanges in the intrapulmonary arteries. In the presentseries the walls of both elastic and muscular preacinararteries were thin and in the elastic arteries the elasticfibres were less dense and less compactly arrangedthan is normal. Other studies have also shown areduction in the amount of elastin, assessed bothhistologically and biochemically, in patients withtetralogy. 18 Such changes probably increase thedistensibility of a vessel wall.

Unlike the preacinar arteries, within the acinus thearteries were significantly smaller than normal. Withthe same quantitative techniques a reduction in intra-

899

acinar arterial size has been reported in other patientswith tetralogy8 and in other conditions of reducedpulmonary blood flow.5 19 201In pulmonary atresia withan intact ventricular septum the intra-acinar arteriesare abnormally small at birth.5 '9 In patients withpulmonary atresia and major aortopulmonarycollateral arteries the intra-acinar arteries inbronchopulmonary segments perfused by smallcollateral arteries are also smaller than normal forage.20 Thus there appears to be a critical level ofpulmonary arterial pressure or blood flow, or both,that is required to ensure normal development of theintra-acinar arteries.Although the intra-acinar arteries were small in

several cases they were increased in number. Anincreased number of intra-acinar arteries has beenreported in other cases of tetralogy.8 On thepostmortem arteriogram the increased density of thebackground haze, the moss-like appearance,' can beattributed both to the dilatation of small preacinararteries and to the increase in number of small intra-acinar arteries.

In the small preacinar arteries and in the intra-acinar arteries the media was thinner than normal inall untreated patients dying at surgery. Previousstudies have given conflicting results. Wagenvoortand Edwards6 reported a reduction in muscularity,while others have found a normal23 or even anincreased wall thickness.8 The thin medial coat isfrequently said to be atrophied in tetralogy, but thisseems unlikely since atrophy implies a reduction inmuscularity, and muscularity is probably nevernormal in postnatal life in patients with tetralogy ofFallot. Indeed babies with pulmonary atresia andintact ventricular septum are born with abnormallythin-walled arteries.5 "' Despite the thinness of themedia, however, it hypertrophies in response to anincrease in pressure, or flow, as was seen after aWaterston-Cooley anastomosis had been performedand in the child with a residual postoperativeventricular septal defect. Recent experimentalstudies have shown that in pulmonary hypertensionsmooth muscle cells increase in size rather than innumber.21 The ability of the medial coat tohypertrophy in tetralogy of Fallot suggests that thechildren are probably born with a normal number ofabnormally small smooth muscle cells in the media.Supporting this hypothesis is the observation that inchildren with tetralogy muscle cells extend as fardown the arterial pathway as in the normal child.The thin-walled, small intra-acinar arteries may

help to explain why after a repair a residual ven-tricular septal defect may be more hazardous thanin a patient with an isolated ventricular defect.Clinically such patients may be difficult to manage,and at necropsy the capillaries are engorged with

copyright. on F



/T


ber 1982. Dow

nloaded from


Johnson, Haworth

blood. The structurally abnormal pulmonary vascularbed may not accommodate the volume of bloodflowing through the residual ventricular septal defectwithout difficulty.The potential complications of shunt procedures

are well recognised.22 23 In case 7 in the present studythe child had had a good shunt in that he improvedclinically and did not develop pulmonaryhypertension. Alveolar development was similar inthe two lungs but unequal perfusion is suggested bythe greater preacinar arterial size and the greater wallthickness of the preacinar and intra-acinar arteries inthe right than in the left lung. Radionuclide studiesafter corrective surgery showed relativehypoperfusion of the lung which had received theshunt, resulting in ventilation-perfusion imbalance.24

In the past the absence of postoperative pulmonaryhypertension has been attributed to a reservecapacity of the pulmonary vascular bed sufficient tocompensate for the reduction in cross-sectional areacaused by thrombosis and intimal change. Suchsevere change was, however, absent in the presentseries and is probably unusual nowadays with earlierpalliative and corrective surgery. Intimal thrombosishas been attributed to a low systemic arterial oxygensaturation (less than 60%), polycythaemia, a lowpulmonary blood flow, or a combination of thesefeatures.2225 We found mild intimal change, how-ever, despite a systemic arterial oxygen saturationgreater than 60%, a haemoglobin concentration of15- 4 g/dl or less, and in one patient a satisfactorypulmonary blood flow after a Waterston-Cooleyanastomosis. Whatever the aetiology of intimalthrombosis, in our patients with a typical andfavourable clinical picture of tetralogy pulmonaryvascular thrombosis could not have caused post-operative pulmonary hypertension.

In several cases the bronchial arteries appearedlarger than normal on the postmortem arteriogramsand on microscopic examination. In the normal lunganastomoses between a preacinar pulmonary arteryand a bronchial artery accompanying the same airwayare uncommon. In our cases of tetralogy examinationof serial sections failed to show an increase in thenumber of such anastomoses. The connectionsnormally found between bronchial arteries andadjacent alveolar duct and alveolar wall arterieswere, however, larger and possibly more frequentthan normal. The normal intra-acinar connectionsbetween the two circulations were also seen.Examination of serial sections confirmed that thedilated arteries at the lung periphery were all dilatedpreacinar pulmonary arteries and not bronchialarteries or dilated anastomosing vessels between thetwo circulations. Thus in life the bronchial arterialsupply probably made an appreciable contribution to

the total pulmonary blood flow but did so throughenlargement of anastomotic channels that are presentin normal lung.

In these patients with tetralogy of Fallot alveolardevelopment was also abnormal. Lung volume wasreduced in relation to both body surface area and age.Other pathological studies have also shown areduction in lung volume in children dying withtetralogy of Fallot.8 In life, a group of adults who hadundergone a successful repair had a reduction inthoracic gas volume to 75% of the predicted value,and the findings were similar in those who had hadinitial palliative surgery and those who had not.26 Themagnitude of the reduction in thoracic gas volume inthis in vivo series is similar to the magnitude of thereduction in lung volume found in the present study.In addition, postoperative lung function studies haveshown only a restrictive defect, there being noevidence of obstruction to air flow.26 In the presentstudy the alveoli and airways appeared normalmacroscopically and microscopically.

Total alveolar number was reduced to a lesserextent than lung volume because the number ofalveoli per unit volume of lung tissue was greater thannormal. In young animals experimental reduction inblood flow is associated with a small lungcomposed ofmany small alveoli.27 These findings are encouragingbecause after corrective surgery the lung probably hasa better chance of achieving normal alveolardevelopment if an increase in the size of existingalveoli rather than the formation of new alveoli isrequired. Performing a corrective operation beforethe critical growth period of the lung is over pre-sumably encourages normal alveolar development.In addition, an early repair permits greater physicalactivity and in normal children physical trainingincreases total lung volume and vital capacity.28 In thelung function study referred to above the mean age atrepair was 19 years and the reduction in thoracic gasvolume in these patients was probably similar to whatit had been in childhood.26Thus in children with tetralogy of Fallot the

pulmonary arterial circulation and the alveoli fail todevelop normally. This should influence the age atwhich the abnormality is corrected because thepotential for growth decreases with age. In thenormal lung the alveoli are formed most rapidlyduring the first three years of life and subsequentlygrowth is by increase in size rather thanxnumber.29 Inthe pulmonary circulation the ability to produceelastin decreases with age and the elastin becomesmore stable, suggesting a reduction in the capacity toremodel as haemodynamic conditions change.30 Theextent to which the intra-acinar arteries can grow andcatch up to achieve a normal size is not known. Thethin-walled muscular arteries probably achieve a

900

copyright. on F



/T


ber 1982. Dow

nloaded from



normal wall thickness since they have the capacity todevelop medial hypertrophy in the presence of toolarge an aortopuhnonary shunt.

In the normal lung growth and remodelliug occurmost rapidly during the first two years of life andrepair of tetralogy of Fallot is probably best carriedout during this time.

This work was supported in part by the British HeartFoundation. We wish to thank the physicians,surgeons, and pathologists of the BromptonHospital, London; Freeman Hospital, Newcastleupon Tyne; and Deutsches Herzzentrum, Munich,West Germany.

References

'Macartney FJ, Taylor JFN, Graham GR, De Leval M,Stark J. The fate of survivors of cardiac surgery ininfancy. Circulation 1980;62:80-91.

2 Damman JF, Ferencz C. The significance of thepulmonary vascular bed in congenital heart disease.I-Normal lungs. II-Malformations of the heart inwhich there is pulmonary stenosis. Am Heart J1956;52:7-17.

Granston AS. Morphologic alterations of the pulmonaryarteries in congenital heart disease. Proc InstMed Chic1958;22: 116-7.

4 Best PV, Heath D. Pulmonary thrombosis in cyanoticcongenital heart disease without pulmonaryhypertension. Journal of Pathology and Bacteriology1958;75:281-91.

Naeye RL. Perinatal changes in the pulmonary vascularbed with stenosis and atresia of the pulmonic valve. AmHeart J 1961;61:586-92.

6 Wagenvoort CA, Edwards JE. The pulmonary arterialtree in pulmonic atresia. Arch Pathol 1961;71:646-53.

7 Wagenvoort CA, Heath D, Edwards JE. The pathologyof the pulmonary vasculature. Springfield, Illinois:Charles C Thomas, 1964.

Hislop A, Reid L. Structural changes in the pulmonaryarteries and veins in tetralogy of Fallot. Br Heart J1973;35: 1178-83.

9 Davies G, Reid L. Growth of the alveoli and pulmonaryarteries in childhood. Thorax 1970;25:669-81.

'0 Hislop A, Reid L. New pathological findings inemphysema of childhood. I-Polyalveolar lobe withemphysema. Thorax 1970;25:682-90.

Dunnill MS. Quantitative methods in the study ofpulmonary pathology. Thorax 1962;17:320-8.

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

Wagenvoort CA, Wagenvoort N. Pathology ofpulmonary hypertension. New York: John Wiley andSons, 1977:48.

4 Brenner 0. Pathology of the vessels of the pulmonary

901

circulation. Part I. Arch Intern Med 1935;56:211-37.5 Gotsman MS, Beck W, Barnard CN, O'Donovan TG,

Schrire V. Results of repair of tetralogy of Fallot.Circulation 1969;40:803-21.

6 Epstein SE, Beiser D, Goldstein RE, Rosing DR,Redwood DR, Morrow AG. Hemodynamicabnormalities in response to mild and intense uprightexercise following operative correction of an atrialseptal defect or tetralogy of Fallot. Circulation1973;47: 1065-75.

Wagenvoort CA, Nauta J, Van der Schaar PJ, WeedaHWH, Wagenvoort N. Vascular changes in pulmonicstenosis and tetralogy of Fallot studied in lung biopsies.Circulation 1967;36:924-32.

18 Farrar JF, Blomfield J, Reye RDK. Structure andcomposition of the pulmonary circulation in congenitalheart disease. Journal of Pathology and Bacteriology1965;90:97-105.

9 Haworth SG, Reid L. Quantitative structural study ofpulmonary circulation in the newborn with pulmonaryatresia. Thorax 1977;32:129-33.

20 Haworth SG, Macartney FJ. Growth and development ofthe pulmonary circulation in pulmonary atresia withventricular septal defect and major aorto-pulmonarycollateral arteries. Br HeartJ 1980;44:14-24.

21 Meyrick B, Reid L. Hypoxia-induced structural changesin the media and adventitia of the rat hilar pulmonaryartery and their regression. Am J Pathol 1980;100: 151-69.

22 Ferencz C. The pulmonary vascular bed in tetralogy ofFallot. II-Changes following a systemic-pulmonaryarterial anastomosis. Bulletin of Johns HopkinsHospital 1960;106:100-18.

23 Newfeld EA, Waldman JD, Paul MH, Muster AJ, ColeRB, Idriss F, Riker W. Pulmonary vascular diseaseafter systemic-pulmonary arterial shunt operations.Am J Cardiol 1977;39:715-26.

24 Alderson PO, Boonvisut S, McKnight RC, Hartman AF.Pulmonary perfusion abnormalities and ventilation-perfusion imbalance in children after total repair oftetralogy of Fallot. Circulation 1976;53:332-7.

25 Heath D, Dushane JW, Wood EH, Edwards JE. Theaetiology of pulmonary thrombosis in cyanoticcongenital heart disease with pulmonary stenosis.Thorax 1958;13:213-7.

26 Bjarke B. Spirometric data, pulmonary ventilation andgas exchange at rest and during exercise in adultpatients with tetralogy of Fallot. Scand J Respir Dis1974;55:47-61.

27 Haworth SG, McKenzie SA, Fitzpatrick ML. Alveolardevelopment following ligation of left pulmonaryartery in newborn pig: clinical relevance to unilateralpulmonary artery. Thorax 1981;36:938-43.

28 Ekblom B. Effect of physical training in adolescent boys.JAppl Physiol 1969;27:350-5.

29 Hislop A, Reid L. Development of the acinus in thehuman lung. Thorax 1974;29:90-4.

30 Farrar JF, Blomfield J, Reye RDK. The structure andcomposition of the maturing pulmonary circulation.Journal ofPathology and Bacteriology 1965;90:83-96.

copyright. on F



/T


ber 1982. Dow

nloaded from


pulmonary vascular and development in tetralogy a correctiontetralogy in patients who usually...

Documents