Br HeartJ 1982; 48: 249-54

Total anomalous pulmonary venous connectionRepair using deep hypothermia and circulatory arrest in 44consecutive infantsD F DICKINSON, K M PARIMELAZHAGAN, M C K TWEEDIE, C R WEST, G P PICCOLI,*F MUSUMECI,* D I HAMILTON

From the Regional Paediatric Cardiothoracic Unit, Royal Liverpool Children's Hospital, Myrtle Street, Liverpool

SUMMARY Forty four consecutive infants aged from 3 days to 10 months underwent repair of totalanomalous pulmonary venous connection using deep hypothermia with circulatory arrest. Therewere eight (18%) early hospital deaths. Using multivariate analysis no significant association couldbe shown between early mortality and age or weight at operation, preoperative pulmonary or

systemic pressure, and preoperative condition for patients undergoing operation during the most

recent five year period. Late pulmonary venous obstruction developed in four (11%) of the survivorsand all of these patients died.

Total anomalous pulmonary venous connection is arelatively uncommon cardiac malformation represent-ing less than 1% of all congenital cardiac anomalies. 1 2Without surgical treatment, the mortality within thefirst year of life is high,3 and those who survive maydevelop progressive pulmonary vascular disease.45Early surgical intervention is therefore indicated inthe majority of patients. Successful surgical correctionwas first reported by Muller in 19516 and subse-quently others described their experience in the surgi-cal management of this lesion, using either hypother-mia with inflow occlusion,7 or standard cardiopulmo-nary bypass.89 It is only with the adoption of deephypothermia and circulatory arrest, however, thatsatisfactory results have been achieved ininfancy. 10- 1 3

In this paper, we review our experience during thelast 12 years of single stage correction of total anomal-ous pulmonary venous connection in infancy usingthis latter technique.

Patients and methods

During the period January 1970 to December 1981,44 consecutive infants under one year of age with totalanomalous pulmonary venous connection underwent

*G P Piccoli and F Musumeci were supported by a grant from the National HeartResearch Fund.

Accepted for publication 13 April 1982

single stage surgical correction at the Royal LiverpoolChildren's Hospital, using deep hypothermia and cir-culatory arrest. Four infants who were operated on in1970 and 1971 using standard cardiopulmonarybypass and previously reported13 are not included inthis report. One further case also included in the pre-vious report,"3 with infracardiac total anomalouspulmonary venous connection occurring as part of acomplex malformation associated with the aspleniasyndrome is excluded since the operation was anattempt at palliation rather than correction.The patients are grouped according to the site of

pulmonary venous connection (Fig. 1). Twenty-twopatients had supracardiac total anomalous pulmonaryvenous connection and nine had infracardiac totalanomalous pulmonary venous connection. The groupof eight patients with intracardiac total anomalouspulmonary venous connection includes six withdrainage entirely to the coronary sinus and two withdrainage direct to the right atrium. Five patients had amixed type of total anomalous pulmonary venousconnection. In four, part of the left upper lobedrained into the innominate vein, while the remainingveins entered the coronary sinus. The fifth patient hadboth upper lobe veins entering the superior vena cavawhile the lower lobe veins entered the right atrium viaa common venous channel.There were 22 male and 22 female patients ranging

in age from 3 to 312 days (mean 77 days) at the time ofoperation. The weight at operation varied from 2 3 kg

249

on Novem

ber 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.48.3.249 on 1 S

eptember 1982. D

ownloaded from

Dickinson, Parimelazhagan, Tweedie, West, Piccoli, Musumeci, Hamilton

10,Supracardiac Cardiac

2 Mixed

Number m L X t Number 5-

1 28 days 29 days-3months 4-12rnoths <0.5 051 099 >1 0

Fig. 1 Age at operation and anatomical type in 44 infants with Pulmonory to systemic pressure ratiototal anomalous pulmonary venous connection. Fig. 2 Preoperative pulmonary to systemic pressure ratio and

anatomical type in 37 infants with total anomalous pulmonaryosrerre-Dfn Z t efH i 1

to 8-7 kg (mean 3-9 kg). One patient was known to

have a ventricular septal defect and the ductusarteriosus was patent in a further 14 patients. Balloonseptostomy was performed at the time of diagnosticcatheterisation in nine patients: four with a supracar-

diac connection, three with drainage to the coronary

sinus, and two with mixed total anomalous pulmonaryvenous connection.

PREOPERATIVE CONDITIONEach patient was assigned retrospectively to one ofthree groups according to the clinical condition at thetime of diagnosis. Group 1 consisted of four infantswhose symptoms were well controlled on medicaltreatment and who underwent operation electively ata mean interval of 122 days (range 95 to 155 days)from the time of presentation. In group 2 there were

27 infants with severe symptoms but no acute deterio-ration in whom operation seemed advisable withoutundue delay. These infants underwent surgery two to54 days (mean 13 days) from the time of diagnosis.Group 3 consisted of 13 critically ill infants whounderwent emergency operation at a mean interval of2-8 days from the time of diagnosis (range less thanone to eight days). Five patients in this group hadsupracardiac and eight had infracardiac total anomal-ous pulmonary venous connection. Eight of thesepatients were ventilated electively before operationbecause of respiratory failure. Critical preoperativestatus was correlated significantly with young age andwith raised pulmonary and lowered systemic pres-sures at the time of cardiac catheterisation (p= 0-05 to0-01 in each case; t tests, confirmed by Mann-Whitney tests).

All cases were operated on using deep hypothermiainduced with core cooling, and total circulatory arrestusing techniques described previously.'3 14 Forpatients with supra- and infracardiac total anomalouspulmonary venous connection. a left anterolateralthoracotomy with transternal extension was used.-5

enous conncLm&&. Aey asJuV ig.

Patients with cardiac or mixed types of total anomal-ous pulmonary venous connection were repairedthrough a median stemotomy. In three most recentpatients with mixed total anomalous pulmonary ven-ous connection predominantly to the coronary sinusthe operation suggested by Van Praagh et al. 16 fortotal anomalous pulmonary venous connection to thecoronary sinus was used.

Preoperative measurements of pulmonary and sys-

temic pressure were made in 37 patients and the pul-monary to systemic pressure ratios for each anatomi-cal group are shown in Fig. 2.

STATISTICAL ANALYSISSeven factors were selected as having a possibleinfluence on early mortality: anatomical type, age andweight at operation, preoperative condition,preoperative Pco2 (without assisted ventilation),preoperative pulmonary and systemic pressures, andthe duration of circulatory arrest. Complete data wereavailable for 29 consecutive patients operated on dur-ing the most recent five year period. These data were

tabulated and the Statistical package for the social sci-ences'7 was used for cross-tabulation, unpaired Stu-dent's t test and discriminant analysis predicting earlydeath or survival. Where necessary, transformation toa logarithmic scale was performed to normalise databefore using Student's t test.

Results

EARLY MORTALITY

Eight infants (18%) died within 30 days of the opera-tion. During the most recent five year period, 29 con-

secutive infants (11 supracardiac, eight infracardiac,seven cardiac, and three mixed) have undergone oper-ation, with four early deaths (14%), compared withfour early deaths among 15 infants (27%) during the

10

250

I

on Novem

ber 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.48.3.249 on 1 S

eptember 1982. D

ownloaded from

Total anomalous pulmonary venous connection

preceding seven years. This difference is not statisti-cally significant. Two of the deaths in the earlierperiod were the result of avoidable technical errors atthe time of operation.

For the 29 patients undergoing operation duringthe most recent five year period, no significant associ-ation could be established between early death andanatomical type, age or weight at operation, preopera-

tive condition, pulmonary or systemic pressures,Pco2, or duration of circulatory arrest. None of thesevariables was significantly different for early deaths as

compared with other cases (unpaired Student's t testconfirmed by Mann-Whitney rank sum test). Nor wasany pattem discernible among the four early deathcases on any two dimensional plot of any pair of con-tinuous variables except in the case of the age/weightplot where they lay very closely on a straight line. Theprobability of early death, computed from a linearfunction on the logit scale by using the generalisedlinear interactive modelling package,'8 did notdepend significantly on any of the above mentionedvariables. Linear discriminant analysis between theearly deaths and the remaining patients usinglogarithmic transformations of age, weight, pulmo-nary and systemic pressure, and duration of circulat-ory arrest was also ineffective in predicting earlydeath. During this period early death appeared to be arandom event not clearly connected with any of thevariables studied, either individually or in combina-tion.

LATE RESULTS

Four patients developed pulmonary venous obstruc-

251

tion, all within three months of the initial operationand all subsequently died. Postoperative cardiaccatheterisation data on these patients are summarisedin the Table. Case 1 underwent a second operationthree months after the initial procedure. The orificesof three of the four pulmonary veins were found to bepartly obstructed by intimal proliferation. An angio-plastic procedure was performed to enlarge the pul-monary veins but the infant died four months later,with clinical evidence of reobstruction of the pulmo-nary veins. Case 2 was readmitted one month after thefirst operation. He deteriorated rapidly after investi-gation and died before a second operation could beperformed. Consent for a necropsy was refused. Case3 had total anomalous pulmonary venous connectionto the coronary sinus. At the second operation theopening between the coronary sinus and the leftatrium was severely stenotic with thickening of thetissue surrounding this ostium and shrinkage of theDacron patch placed at the mouth of the coronary

sinus. The original operation was revised but thepatient died suddenly two months later. At necropsythe inlet of the coronary sinus to the left atrium waswidely patent, but the four pulmonary veins enteredthe apex of the coronary sinus via a single stenoticorifice of less than 4 mm diameter. Case 4 had drain-age of the pulmonary veins direct to the right atrium.He died before a second operation could be performedand at necropsy it was found that the right pulmonaryveins were totally occluded by shrinkage and thicken-ing of the original Dacron baffle used to redirect pul-monary venous blood into the left atrium.The 32 longer term survivors have been reviewed

Table Postoperative cardiac catheterisation data in 11 patients

Case No. Anatomical Interval after Pressures (mmHg) Pulmonary to Commenttype operation systemicflow ratio

(mth) Pulmonary Systemic1 Supracardiac 3 100 65 1 Pulmonary venous obstruction;

reoperation; died2 Infracardiac 1 65 120 1 Pulmonary venous obstruction;

died3 Cardiac 3 80 75 1 Pulmonary venous obstruction;

reoperation; died4 Cardiac 2 90 75 1 Pulmonary venous obstruction;

died5 Supracardiac 59 21 100 2 Residual ASD;

successful operation6 Supracardiac 16 78 84 1-8 Residual VSD;

successful operation7 Infracardiac 1 35 80 1 Transient pericardial

effusion; good result8 Cardiac 13 36 80 1 Good haemodynaniic result;

atrioventricular block9 Cardiac 12 28 100 1 Good result10 Cardiac 26 23 110 1 Good result11 Mixed 3 28 95 2-2 Residual ASD;

awaiting reoperation

ASD, atrial septal defect; VSD, ventricular septal defect.

on Novem

ber 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.48.3.249 on 1 S

eptember 1982. D

ownloaded from

Dickinson, Parimelazhagan, Tweedie, West, Piccoli, Musumeci, Hamilton

regularly with a follow up ranging from three monthsto 10 years (mean 5*1 years). One child (case 5, Table)required a second operation to close a residual atrialseptal defect and a second (case 6, Table) has had asecond operation to close a ventricular septal defect. Afurther five patients have been investigated postopera-tively (Table). Four have a good haemodynamic resultbut one has atrioventricular block and one had a trans-ient pericardial effusion at the time of investigation.Two of these patients were investigated because of aslight suspicion on clinical grounds that pulmonaryvenous obstruction was developing but these suspi-cions were unfounded. Pulmonary venous obstructionhas not been identified in any patient as a later post-operative complication. The last patient (case 11,Table) has a residual atrial shunt after repair of amixed type of total anomalous pulmonary venousconnection and will need a second operation in thefuture.The remaining patients are clinically well with no

abnormal signs in the cardiovascular system otherthan wide splitting of the second heart sound and mildcardiomegaly on chest x-ray.

Discussion

Before 1970, the surgical mortality for correction oftotal anomalous pulmonary venous connection ininfancy was significantly higher than in older patientsand some authors regarded young age as a major fac-tor in determining early mortality. 1 9 During the pastdecade, improved results have been achieved particu-larly in infants under 1 year of age. 11-13 20-22 In theseseries, young age at operation has not been confirmedas a risk factor and our present results substantiatethis view. During the 12 year period reviewed, therewere no deaths among five patients operated on dur-ing the first week of life showing that even theyoungest patients can be corrected successfully.Though we were unable to show a correlation betweenpreoperative condition and early mortality this is onlylikely to be the case if the operation is performed withan appropriate degree of urgency. We agree with theview expressed by many authors1' 122021 that anydelay in the effective surgical treatment of symptoma-tic infants, particularly when pulmonary venousreturn is obstructed, may result in a higher operativemortality rate.

Surgical mortality in the patients reported byClarke et al.12 and Turley et al.20 was most closelyrelated to the anatomical type of total anomalous pul-monary venous connection present. During our mostrecent experience no such correlation was presentwith survival of all three patients with mixed totalanomalous pulmonary venous connection. It should

be noted, however, that these patients had a relativelysimple type of mixed total anomalous pulmonary ven-ous connection predominantly to the coronary sinus.In more complex types a higher mortality might beexpected.2' It is interesting that the anatomical typecarrying the highest mortality in the series of Clarkeetal..'2 Turley et al.,20 and Hammon et al. 22 (infracar-diac) carried the lowest mortality among our patients.Conversely, we have experienced a relatively highearly mortality among patients with the intracardiactype of the anomaly, a defect which in otherhands"' 1220 carries a low mortality. The number ofdeaths in each series, however, is small and manydeaths were related to factors individual to eachpatient rather than specific for a particular anatomicaltype. For example, two deaths in the present seriesand two deaths among the patients reported byWhight et al. I I were mainly the result of avoidabletechnical errors unrelated to the particular anatomicaltype present. Reviewing the combined recent experi-ence of the groups in Auckland, Boston, andBirmingham, Katz et al. 21 concluded that anatomicaltype was not a determinant of early mortality exceptin the case of patients with the complicated mixedtype of total anomalous pulmonary venous connec-tion. Our results show that an early mortality of lessthan 15% should be expected in patients with thesupra- and infracardiac types. Others" 1220 haveshown that a similar or lower mortality can beachieved with total anomalous pulmonary venousconnection to the right atrium or coronary sinus.Though late mortality after correction of total

anomalous pulmonary venous connection is low,23 inmany recently reported series there has beensignificant mortality and morbidity associated withthe development of pulmonary venous obstruc-tion. "I 20 24 Four (11%) of the 36 hospital survivors inour series developed this complication, all withinthree months of the original operation. In one case,the main reason for obstruction appeared to be relatedto shrinkage and thickening of a synthetic Dacronpatch, but in two cases, one with supracardiac drain-age and one with the coronary sinus type, obstructiondeveloped at the orifices of the pulmonary veins at asite remote from any suture line. The cause of obs-truction in the fourth case was not determined.Though pulmonary venous obstruction has beendescribed in patients with total anomalous pulmonaryvenous connection before surgical treatment,3 therewas no evidence of this in any of our patients. Similarfindings have been reported by other authors. BothTurley et al.20 and Fleming et al.24 reported cases ofinfradiaphragmatic total anomalous pulmonary ven-ous connection developing pulmonary venous obs-truction at a site proximal to the original anastomosis.In the study of Whight et al. " three patients with

252

on Novem

ber 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.48.3.249 on 1 S

eptember 1982. D

ownloaded from

Total anomalous pulmonary venous connection

total anomalous pulmonary venous connection to thecoronary sinus developed pulmonary venous obstruc-tion, caused in one case by thickening of a pericardialpatch and in two further cases by partial obliterationof the ostia of the pulmonary veins by fibrosis. Surgi-cal treatment of pulmonary vein obstruction isdifficult"l20 and the problem may recur even afterapparently successful reoperation.20 Its occurrence isunpredictable and though it appears to be more com-mon after repair of the infradiaphragmatic type, it isby no means confined to that group.

Several approaches have been used to gain optimalexposure of the anomalous venous channels in supra-and infracardiac total anomalous pulmonary venousconnection. Throughout the 12 year period, we haveused the left anterolateral thoracotomy advocated byRoe.'5 It has been suggested that aortic cannulationmight prove to be difficult using this approach andthat cannulation of the descending aorta might benecessary.'5 With trans-sternal extension of the inci-sion, however, we have not found this to be the case.In our experience, the left anterolateral thoracotomygives excellent exposure of the left atrium and com-mon pulmonary vein and facilitates the constructionof the largest possible anastomosis in spite of the smallsize of the left atrium in this anomaly.25 26 If neces-sary, the incision can be extended onto the left atrialappendage which is usually of relatively normal size.26We have not found it necessary during this 12 yearperiod to enlarge the size of the atrial cavity in anypatient either by using a patch or by moving theinteratrial septum to the right. No problems withkinking of the anastomosis have been encounteredand pulmonary venous obstruction at the suture linehas not been recognised as a late complication in anypatient with these anatomical types of total anomalouspulmonary venous connection.

Several authors"l2225 have published their findingsat cardiac catheterisation carried out one to six yearsafter operation for total anomalous pulmonary venousconnection. Their findings show normal or near nor-mal right and left ventricular and left atrial volumes,with normal right ventricular pressures in the ma-jority of asymptomatic survivors. Unexpected majorresidual defects were not identified. In view of theseencouraging reports, we have studied our postopera-tive patients only when specific clinical indicationswere present.

Conventional repair of total anomalous pulmonaryvenous connection to the coronary sinus involvesincising the margin of the coronary sinus and carryingthe incision back through the foramen ovale into theleft atrium. A large patch is then used to roof over thecombined orifice of the coronary sinus and fossaovale. An alternative procedure was suggested by VanPraagh et al. 16 in 1972. Though the procedure was

253

originally devised to avoid damage to internodal con-duction pathways, its main advantage appears to bethe avoidance of prosthetic material within the atriumof a small infant. In view of the problems encounteredafter patch shrinkage both in our patients and thosereported by Whight et al., .this advantage appears tobe considerable. We have used this method in threerecent cases with mixed type total anomalous pulmo-nary venous connection predominantly to the coro-nary sinus and intend to use it more extensively in thefuture.The role of balloon atrial septostomy in the man-

agement of the infant with total anomalous pulmonaryvenous connection remains a subject for debate. Leftheart filling in this condition is entirely dependent ona right to left atrial shunt, and the atrial septum maybe a site of obstruction even in patients with no pres-sure gradients between left and right atrium.26 Bal-loon septostomy has therefore been recommended forall infants with total anomalous pulmonary venousconnection as a means of palliating the symptomaticinfant.27 28 Patients with pulmonary venous obstruc-tion are unlikely to be benefited. The procedure,however, may be helpful in infants in whom the pul-monary venous return is unobstructed, pulmonaryartery pressure is less than half systemic, and post-ponement of operation for a short period is desired.Whight et al. "I have drawn attention to a possibledisadvantage of balloon septostomy in patients withtotal anomalous pulmonary venous connection to thecoronary sinus in whom the Van Praagh technique'6is to be used for repair, in that closure of the atrialseptum may be easier if the fossa ovale is intact. Wehave used balloon atrial septostomy in nine patients inan attempt to improve left heart filling preoperatively.Though it has been suggested that septostomy may bedifficult in total anomalous pulmonary venous con-nection,27 using biplane screening facilities, the pro-cedure has not been accompanied by any morbidity inour patients. We do not, however, regard this proce-dure as a substitute for urgent surgical treatment inthe majority of patients.Our experience with the management of infants

with total anomalous pulmonary venous connectionduring the past 12 years suggests that the best resultsare obtained when referral to a specialised paediatriccardiac unit is made as soon as the diagnosis is sus-pected, preferably during the first weeks of life. Sincesurgical mortality is not correlated with young age,operation should be performed on an emergency orsemi-emergency basis as the patient's conditiondemands. Unnecessary delay is likely to lead to acutedeterioration particularly when pulmonary venousobstruction is present. The long term results andexcellent health of the surviving children are justreward to the medical and surgical team who accept

on Novem

ber 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.48.3.249 on 1 S

eptember 1982. D

ownloaded from

Dickinson, Parimelazhagan, Tweedie, West, Piccoli, Musumeci, Hamilton

the considerable challenge which these seriously illinfants present on admission to hospital.

References

1 Hoffman JIE, Christianson R. Congenital heart diseasein a cohort of 19,502 births with long-term follow up. AmJ Cardiol 1978; 42: 641-7.

2 Dickinson DF, Arnold R, Wilkinson JL. Congenitalheart disease among 160,480 liveborn children in Liver-pool 1960 to 1%9. The implications for surgical treat-ment. Br HeartJ 1981; 46: 55-62.

3 Gathman GE, Nadas AS. Total anomalous pulmonaryvenous connection. Clinical and physiologic observationsof 75 pediatric patients. Circulation 1970; 42: 143-54.

4 Haworth SG, Reid L. Structural study of pulmonary cir-culation and of heart in total anomalous pulmonary ven-ous return in early infancy. Br HeartJ 1979; 39: 80-92.

S Newfeld EA, Wilson A, Paul MH, Reisch JS. Pulmonaryvascular disease in total anomalous pulmonary venousdrainage. Circulation 1980; 61: 103-9.

6 Muller WH Jr. The surgical treatment of transposition ofthe pulmonary veins. Ann Surg 1951; 134: 683-93.

7 Lewis FJ, Varco RL, Taufic M, Niazi SA. Direct visionrepair of triatrial heart and total anomalous pulmonaryvenous drainage. Surg Gynecol Obstet 1956; 102: 713-20.

8 Cooley DA, Ochsner A Jr. Correction of total anomalouspulmonary venous drainage. Surgery 1957; 42: 1014-21.

9 Cooley DA, Hallman GL, Leachman RD. Total anomal-ous pulmonary venous drainage: correction with the useof cardiopulmonary bypass in 62 cases. J Thorac Car-diovasc Surg 1966; 51: 88-102.

10 Dillard DH, Mohri H, Hessel EA, et al. Correction oftotal anomalous pulmonary venous drainage in infancyutilizing deep hypothermia with total circulatory arrest.Circulation 1967; 35, suppl 1: 105-10.

11 Whight CM, Barratt-Boyes BG, Calder AL, Neutze JM,Brandt PWT. Total anomalous pulmonary venous con-nection. Long-term results following repair in infancy. JThorac Cardiovasc Surg 1978; 75: 52-63.

12 Clarke DR, Stark J, De Leval M, Pincott JR, TaylorJFN. Total anomalous pulmonary venous drainage ininfancy. Br Heart J 1977; 39: 436-44.

13 Di Eusanio G, Sandrasagra FA, Donnelly RJ, HamiltonDI. Total anomalous pulmonary venous connection (sur-gical technique, early and late results). Thorax 1978; 33:275-82.

14 Hamilton DI, Shackleton J, Rees GJ, Abbott T. Experi-ence with deep hypothermia in infancy using core cool-ing. In: Barratt-Boyes BG, Neutze JM, Harris EA, eds.Heart disease in infancy. Edinburgh and London: Chur-chill Livingstone, 1973: 52-74.

15 Roe BB. Posterior approach to correction of total

anomalous pulmonary venous return. J Thorac Cardiuvasc Surg 1970; 59: 748-53.

16 Van Praagh R, Harken AH, Delisle G, Ando M, GrossRE. Total anomalous pulmonary venous drainage to thecoronary sinus. J Thorac Cardiovasc Surg 1972; 64: 132-5.

17 Nie NH, Hull CH, Jenkins JG, Steinbrenner K, BentBH. Statisticalpackagefor the social sciences. 2nd ed. NewYork: McGraw-Hill, 1975.

18 Baker RJ, Nelder JA. The GLIM system. Release 3.Oxford: Numerical Algorithms Group, 1978.

19 Wukasch DC, Deutsch M, Reul GJ, Hallman GL,Cooley DA. Total anomalous pulmonary venous return:review of 125 patients treated surgically. Ann Thorac Surg1975; 19: 622-33.

20 Turley K, Tucker- WY, Ullyot DJ, Ebert PA. Totalanomalous pulmonary venous connection in infancy:influence of age and type of lesion. Am J Cardiol 1980;45: 92-7.

21 Katz NM, Kirklin JW, Pacifico AD. Concepts and prac-tices in surgery for total anomalous pulmonary connec-tion. Ann Thorqc Surg 1978; 25: 479-87.

22 Hammon JW Jr, BenderHW Jr, Graham TP Jr, BoucekRJ Jr, Smith CW, Erath HG Jr. Total anomalous pul-monary venous connection in infancy. Ten years' experi-ence including studies of postoperative ventricular func-tion. J Thorac Cardiovasc Swrg 1980; 80: 544-51.

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

24 Fleming WH, Clark EB, Dooley KJ, et al. Late compli-cations following surgical repair of total anomalous pul-monary venous return below the diaphragm. Ann Thor"cSurg 1979; 27: 435-9.

25 Graham TP Jr, Jarmakani JM, Canent RV Jr. Left heartvolume characteristics with a right ventricular volumeoverload: total anomalous pulmonary venous connectionand large atrial septal defect. Circulation 1972; 45: 389-96.

26 Mathew R, Thilenius OG, Replogle RL, Arcilla RA.Cardiac function in total anomalous pulmonary venousreturn before and after surgery. Circulaion 1977; 55:361-70.

27 El-Said G, Mullins CE, McNamara DG. Management oftotal anomalous pulmonary venous return. Circulation1971; 45: 1240-S50.

28 Mullins CE, El-Said GM, Neches WH, et al. Ballon atrialseptostomy for total anomalous pulmonary venousreturn. Br Heart J 1973; 35: 752-7.

Requests for reprints to Mr DI Hamilton, RoyalLiverpool Children's Hospital, Myrtle Street, Liver-pool L7 7DG.

254

on Novem

ber 30, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.48.3.249 on 1 S

eptember 1982. D

ownloaded from