Surgical management of multiple ventricular septal defects

Volume 80, Number 4

October 1980

THORACIC AND CARDIOVASCULAR SURGERY The Journal of

J

THORAC CARDIOVASC SURG

80:485-493, 1980

Original Communications

Surgical management of multiple ventricular septal defects Between Jan. I. 1973, and June I, 1979, 29 patients underwent surgical closure of multiple ventricular septal defects (VSDs). Ineluded are 19 patients with VSDs only (one death), five with tetralogy of Fallot (TF) (0 deaths), two with transposition of the great arteries and VSD (TGAIVSD) (one death), two with complete atrioventicular (AV) canal (one death), and one with common atrium (one death). The overall hospital mortality rate was 14% (4/29) and was significantly related to the complexity of the underlying malformation (p = 0.01) and the presence of major associated cardiac lesions (p = 0.005). The incidence of reoperation for overlooked VSDs was 28% (8/29) and was significantly related to the presence of muscular VSDs (all overlooked defects were muscular) and the failure preoperatively to diagnose multiple VSDs (p = 0.05). Of patients with shunt data available after the final operation, 85% (17/20) had a QI'/Qs :s 1.5, and no patient had a final Qp/Qs > 2.0. This experience indicates that satisfactory results usually can be achieved after operation for multiple VSDs and emphasizes the value of early assessment for residual shunting and reoperation when indicated.

James K. Kirklin, M.D., Aldo R. Castaneda, M.D., John F. Keane, M.D., Kenneth E. Fellows, M.D., and William I. Norwood, M.D., Boston, Mass.

Diagnostic techniques, \, 2 operative repair,":" and surgical results7-13 are well established for patients with a single large ventricular septal defect (VSD). In patients with multiple VSDs, diagnosis is less seFrom the Departments of Cardiovascular Surgery, Cardiology, and Radiology, Children's Hospital Medical Center and Harvard Medical School, Boston, Mass. Received for publication Feb. 26, 1980. Accepted for publication April 22, 1980. Address for reprints: Aldo R. Castaneda, M.D., Department of Cardiovascular Surgery, Children's Hospital Medical Center, 300 Longwood Ave., Boston, Mass. 02115.

cure, 14-17 surgical techniques are more varied,": 18 and published mortality rates are higher than in single VSD.14, 15 Regardless of underlying disease, all patients with multiple VSDs have important features in common. This report describes our experience with surgical repair of multiple VSDs since 1973, both as isolated lesions and in association with other cardiac malformations. Material and methods Patients. Between Jan. 1, 1973, and June 1, 1979, 29 patients with multiple VSDs underwent surgical repair. Nineteen patients had VSDs only, five had tetral-

0022-5223/80/100485+09$00.90/0 © 1980 The C. V. Mosby Co.

485

The Journal of

486

Kirklin et at.

Thoracic and Cardiovascular Surgery

2

/

3

4 b

~

- r ----Cl.d Fig. 1. Anatomic classification of ventricular septal defects. J, Infundibulo-ventricular (perimembranous). 2,

Subpulmonary (intrainfundibular). 3, Atrioventricular canal type. 4, Muscular VSDs which include (a) anterior, mid-muscular, (c) posterior (infratricuspid or inlet septal), and (d) apical.

(b)

ogy of Fallot (TF), two had transposition of the great arteries with VSD (TGA/VSD), two had complete atrioventricular canal defect (A V canal), and one patient had common atrium and cleft mitral valve with situs ambiguus and polysplenia. Age at the time of first repair ranged from 5 weeks to 9 years with a mean of 3.0 years and a median of 20 months. Nine of the 29 patients had previously placed pulmonary artery bands. Six of the nine had VSDs only, two had complete A V canal, and one had TGA/VSD. Major associated cardiac lesions were present in nine of the 29 patients. These included patent ductus arteriosus (four patients), pulmonary stenosis (two patients), coarctation of the aorta, aortic stenosis, subaortic stenosis, double mitral valve orifice, mitral regurgitation, right pulmonary artery stenosis, situs ambiguus with interrupted inferior vena cava, and congenital heart block (one patient each). Anatomic classification. VSDs were categorized by location into four groups: infundibulo-ventricular (perimembranous), subpulmonary (intrainfundibular),

AV canal type, and muscular (Fig. I). Muscular VSDs were further subdivided to include anterior, mid, posterior (infratricuspid or inlet septal), and apical muscular defects. Since precise counting of multiple muscular defects is not always possible, we have numbered simply as "numerous" three or more discrete holes in anyone portion of the muscular septum (anterior, mid, posterior, or apical). Most (42 of 67) of the VSDs were in the muscular septum, but 20 were infundibulo-ventricular, four were A V canal type, and the remaining defect was subpulmonary (Table I). The most common combinations of VSDs were infundibulo-ventricular plus muscular defects, and all but three patients had one or more muscular defects (Table II). Operative technique. Standard cardiopulmonary bypass techniques were used for patients weighing 8 kg or more, and deep hypothermia and circulatory arrest (with either surface and core cooling or more recently, core cooling alone) for those under 8 kg. Cold cardioplegia was used in some cases. Surgical exposure of the VSDs was through the right

Volume 80

48 7

Surgical management of multiple VSDs

Number 4 October, 1980

Table I. VSD types in each malformation Malformation VSDtype

VSDs only

lnf-vent Subpulmonary AVC type Muscular Anterior Mid Posterior Apical Numerous" Anterior Mid Posterior Apical

13

Totals

47

I

TF

I

I

TGAIVSD

5

AVC

I

Common atrium

2

I 2

20 1 4

2

5 5

8 6 3

2

2 13

1 2

2

Total

17

8 I 2 I 2 10

4

1

42

67

5

Legend: VSD, Ventricular septal defect. TF, Tetralogy of Fallot. TGA/VSD, Transposition of the great arteries with VSD. AVC, Common atrioventricular

canal, lnf-vent, Infundibulo-ventricular. "Numerous (three or more) defects located in any portion of the muscular septum.

atrium, right ventricle, or left ventricle. Large defects or groups of small muscular defects were closed by placement of a Dacron patch, whereas isolated small muscular defects were closed with mattress sutures reinforced with felt pledgets. Calculation of postoperative shunts. Samples were obtained from catheters placed at operation in the right atrium, pulmonary artery, and radial artery for determination of oxygen saturations. Qp/Qs was calculated by means of the standard shunt approximation: Ql'/Qs

== (Sat ao, - Sat vo,)/(Sat PVo2

-

Sat PAo2 ) .

Pulmonary arterial, systemic arterial, and right atrial samples were obtained directly, and pulmonary venous oxygen saturation was assumed equal to systemic arterial saturation. Inspired oxygen concentration was set at approximately 40% during the measurements. Shunt calculations obtained during cardiac catheterization were calculated with oxygen content in the usual manner. Results Incidence. During this time period, multiple VSD has occurred in approximately 12% of patients with VSD only, 2% of patients with TF, 2% of patients with TGAIVSD, and 2% of patients with A V canal operated upon at our institution. Diagnosis. The diagnosis of multiple VSDs was first made preoperatively by cineangiogram in 16 patients (55%), intraoperatively in seven patients (24%), postoperatively in five patients (17%), and at autopsy in one patient (3%) (Table III).

Table II. Combinations of defects in each malformation Malformation VSDs only

TF TGA/VSD AVC Common atrium Total

Types of defects Inf-vent + A VC type Inf-vent + subpulmonary Inf-vent + muscular" Musculart Inf-vent + muscular" Inf-vent + muscular" A VC type + muscular" Musculart

No. of patients

2 I 10

6 5 2 2 I

29

For legend see Table I. "One or more defects. t Two or more defects.

Approach for VSD closure. A transatrial approach was used for 93% (14/15) of nontetralogy infundibulo-ventricular VSDs, all (4/4) AV canal type VSDs, 89% (8/9) of mid-muscular defects, and all (4/4) posterior muscular VSDs (Table IV). A right ventriculotomy was used for all (SIS) infundibulo-ventricular defects in TF, for the one patient with a subpulmonary defect, and for 78% (7/9) of anterior muscular VSDs. A left ventriculotomy was used for closure of 63% (12/19) of apical muscular VSDs. Multiple approaches at a single operation were used in 29% (11/38) of operations for VSD closure (Table V). Combined right and left ventriculotomies were used in 11% (4/38) of operations. Reoperations for VSD closure. Twenty-eight per-

488

The Journal of Thoracic and Cardiovascular Surgery

Kirklin et al.

Table III. Timing of multiple VSD diagnosis Time of diagnosis of multiple VSDs Preoperative Intraoperative Postoperative Earlyt

Later

Malformation VSDs only 12

5 (1)*

I

I

TF

Total

I

TGA/VSD

AVC

I

Common atrium

2 (I)

2 I

I (I)

2

0 I (I)

Autopsy§ Totals

5

19

2

2

No.

1

%

16

55

7

24

4 I I

14

3 3

29

For legend see Table I. 'Hospital deaths in parentheses. t Less than I week after original operation. :j:Twomonths or more after original operation. §Single mid-muscular defect repaired at operation; numerous mid-muscular defects seen at autopsy.

Table IV. Route of VSD closure

VSD type Inf-vent

Subpulmonary

AVe type

Anterior muscular

Mid-muscular

Posterior muscular

Apical muscular

Totals

Approach for closure RA RV LV RA RV LV RA RV LV RA RV LV RA RV LV RA RV LV RA RV LV

Malformation VSDs only

I

TF

12 I

I

TGAIVSD

I

AVC

2

I

Common atrium

Total * 14 6

5

0 I

0 2

2

4

0 0 I 5

2

6t

7 0 8 I 0

3

4

2

0 0 7 0

7 8

2

46

10

4

2

12

5

66t

Legend: RA, Right atrium. RV, Right ventricular. LV, Left ventricle. For other abbreviations see Table I.

'This includes only VSDs actually closed; recurrent VSO's which were previously closed are not recounted. Oefects listed as "numerous muscular" in Table II are listed as one defect if closed with one patch. tTotal differs from that in Table II since one patient had numerous mid-muscular defects, which were not closed at final operation.

cent (8/29) of patients underwent reoperation for closure of residual VSDs (Table VI). Of these, six patients had two operations and two patients had three operations for VSD closure. Of the eight patients requiring reoperation, six patients underwent reoperation within I week of the original operation (five of these within 24 hours of the

original operation). All patients undergoing early reoperation had a calculated Qp/Qs of 2.0 or more before reoperation. Of the six patients having early reoperation, three underwent cardiac catheterization in the interval between operations, and the residual defects were localized successfully. Three patients had late reoperation, 2 months or

Volume 80

Surgical management of multiple VSDs

Number 4

48 9

October, 1980

Table V. Approaches for VSD closure at a single operation Operative approach

Malformation VSDs only

RA RV LV RA, RV RA,LV RV,LV RA, RV, and LV

10 (I)t 2 I 2 4

Totals

21

1

TF

\

I

TGAIVSD 3 (I)

AVC

I

Common atrium

Total *

I (I)

17 (3) 9 1 2 5 (I) 2 2

3

7

I (I) 2

2 9

3

38 (4)

4

For legend see Tables I and I V. 'No. of operations performed, not the number of patients. One patient underwent early re-exploration without identification of additional defects and is not included here. tDeaths indicated by parentheses.

more after the original procedure. In addition to residual VSDs, important right ventricular outflow obstruction after repair of TF was present in two of these patients. All overlooked VSDs were muscular defects (Table VII). Anterior muscular defects were most frequently overlooked at the original operation, 56% (5/9) necessitating reoperation for closure. Thirty percent (3/10) of mid-muscular, 25% (1/4) of posterior muscular, and 16% (3/19) of apical muscular defects were overlooked at the initial operation. One infundibulo-ventricular defect was reclosed at reoperation after having been patched previously. Age at operation, type of malformation, presence of major associated lesions, presence of previous pulmonary artery banding, and number of VSDs had no significant relationship to the incidence of reoperation (p > 0.2). The type of VSD was significant in that all overlooked VSDs were muscular defects, and the relationship of anterior muscular VSD to reoperation was probably significant when compared to the relationship of all other muscular VSDs to reoperation (p = 0.06). A significant relationship existed between the absence of a preoperative diagnosis of multiple VSD and the incidence of reoperation (p = 0.05) (Table VIII). Hospital mortality rate. The overall hospital mortality rate in this group of patients with multiple VSDs was 14% (4/29) (Table IX). The mode of death in all patients was low cardiac output. The mortality rate was highest in patients less than 3 months of age and lowest in patients over 2 years of age. Mortality rate by malformation was lowest in VSDs only and TF. No relationship could be demonstrated between hospital mortality rate and time of diagnosis, presence of previous pulmonary artery banding, types of VSD

Table VI. Number of VSD operations per patient (op./pt.) No. op./pt. I 2 3 Totals

Time of reoperation*

Early Late Early/late Late/late

Mortality

No.

I I %

70% CL

21

3t

14

6%-27%

8

It

12

7%-24%

4

14

7%-24%

No. of patients

5(1)1 )

I I I

29

Legend: CL, Confidence limits. 'Early = less than I week after the original operation. Late = 2 months or

more after the original operation. EarlyIlate = first reoperation early and second reoperation late. Latellate = second reoperation was 1 day after first reoperation. tThe p value for difference = 0.70. ~Hospital death in parentheses.

present, route of closure of VSD, multiple ventriculotomies, early versus late reoperation, or number of operations for VSD closure (p > 0.2). By combining VSDs only and TF together as "simple malformations," and TGA/VSD, AV canal, and common atrium together as "complex" malformations, we find that a significant difference existed between mortality rates in the two groups-one death in 24 patients versus three deaths in five patients, p = 0.01 (Table IX). Analysis of the relationship between age at operation and mortality rate is limited by the small number of patients, but a significant difference did exist between patients 6 months of age or less at initial operation (three deaths in eight patients) and patients over 6 months of age (one death in 21 patients) (p = 0.05) (Table IX).

The Journal of

490

Kirklinetal.

Thoracic and Cardiovascular Surgery

Table VII. Reoperationfor residual VSD Reoperations for VSD closure Early *

VSD location

Late t

o o o

Inf-vent Subpulmonary AVe type Anterior muscular Mid-muscular] Posterior muscular Apical muscular

l:j:

o o o

5 (l)§ 1 1

1 l:j:

2

1

Totals

9 VSDs (6 operations)

4 VSDs (4 operations)

Mortality

1/6 = 17%

0/4 = 0%

(2%-46%)~

(0%-38%)~

= 0.60. For abbreviations see Table I. 'Less than I week after original operation. tTwo months or more after original operation. Recurrence of previously closed YSD. § Deaths in parentheses. IIOne patient had numerous overlooked mid-muscular defects noted at autopsy and is not included in this table. ~ Percentages in parentheses indicate 70% confidence limits. Legend: The p value for difference (Fisher)

*

Table VIII. Timing of diagnosis of multiple VSDs versus No. of operations No. of operations More than one

One

Time of diagnosis

No.

Preoperati ve Intra. or postop.

14 7

I

% of total

No.

88

2

54

6

12* 46*

Total 16 13

29

Total 'The p value for difference (Fisher)

I % of total

=

0.05.

A significant relationship was demonstrated between the presence of major associated cardiac lesions and mortality rate, four deaths occurring in nine patients (44%) with major associated lesions and no deaths in 20 patients (0%) without major associated lesions (p value for difference = 0.005). When the patients with major associated lesions were further divided according to age, the relationship between age (less than 6 mo) and mortality rate was possibly significant (p = 0.17) (Table X). Postoperative rhythm. At the time of hospital discharge, 83% (15/18) of survivors with VSDs only and all survivors with TF or A V canal were in normal sinus rhythm. One patient with A V canal had complete heart block at the time of death, and the patient with common atrium had congenital complete heart block. Among patients with VSDs only, one patient required a permanent pacemaker for sick sinus syndrome, and two pa-

tients were discharged with an ectopic atrial rhythm. One patient with TGA/ VSD was discharged with a junctional rhythm. Final status of ventricular septum. Twenty of the 25 hospital survivors had shunt calculations available after the final operation (Table XI). Sixteen patients had in-hospital shunt calculations and eight patients had late cardiac catheterizations. Both in-hospital and late shunt data were available in four patients. In two of these patients the Qp/Qs was equal to 1.0 at both studies, in one the Qp/Qs was 1.2 early and 1.0 late, and in one the Qp/Qs was 1.6 early and 1.4 late. Thus of the 20 patients with shunt data available after the last procedure, eight (40%) are known to have no shunt, nine (45%) to have a small shunt, three (15%) to have a moderate shunt, and none to have a large shunt. The chances of attaining a Qp/Qs ~ 2.0 after each operation were about 70% (Table XII). Discussion Although the incidence of multiple VSDs is low, the associated diagnostic and technical problems make it an important lesion. Although our experience only includes multiple VSDs in association with VSDs only, TF, TGA/VSD, AV canal, and common atrium, the lesion also occurs with double-outlet right ventricle.'? and, undoubtedly, many other malformations. Diagnosis. The importance of preoperative angiography is demonstrated by the significant correlation between the failure preoperatively to diagnose multiple VSDs and reoperation for overlooked VSDs. Accurate

Volume 80

Surgical management of multiple VSDs

Number 4

49 I

October, 1980

Table IX. Deaths by age and malformation at Children's Hospital Medical Center (January, 1973, to June, 1979) Malformation Age (mo) <;3 >3<;6 >6 <;12 >12<;24 >24 Totals Hospital deaths 70% CL

I

VSDs only

TF

1 (1)* 3* 4t 2 9t 19 (I)

I

TGA/VSD

I

I

AVC

Common atrium

No.

I I %

70% CL

3 5 4 4 13 29

2 I 0 1 0 4

67 20 0 25 0 14

24%-96%} :j: 3%-53%

1 (1)* 1(1)* I 3 5 (0)

5% (1%-17%)

0% (0%-32%)

1(1)* 1 2 (I) 50% (7%-93%)

1 (I)

2 (I) 50% (7%-93%)

§

Hospital deaths

No. of patients

0%-38%} 3%-62% :j: 0%-14% 7%-24%

100% (15%-100%)

§

Legend: Figures in parentheses are number of deaths. For abbreviations see Table I. *Patient with major associated lesions. t Two patients with major associated lesions.

t The p value for difference (Fisher) = 0.05. § The p value for difference (Fisher)

= 0.0 I.

localization of all defects in a given patient is difficult but should be possible more frequently with current angled-oblique views.t" The importance of intraoperative diagnostic maneuvers is indicated by the high incidence (25%) of intraoperative diagnosis. Operative approach. The data presented here indicate that a transatrial approach is generally satisfactory for nontetralogy infundibulo-ventricular defects, A V canal type, mid-muscular, and posterior muscular VSDs. Satisfactory results can be achieved with a right ventriculotomy for infundibulo-ventricular defects in TF, subpulmonary defects, and anterior muscular VSDs. Apical muscular VSDs occasionally can be well visualized through the right atrium or right ventricle, but a left ventriculotomy is frequently necessary for accurate visualization and closure of the defect. 21-24 Combined right and left ventriculotomy may be employed when necessary without a demonstrated increase in risk. Reoperations for VSD closure. The difficulties in accurate intraoperative assessment of the ventricular septum are emphasized by the important incidence (28%) of reoperation for residual defects, most of which were overlooked at the initial operation. The surgeon must have a systematic method of evaluating the ventricular septum and a knowledge of combinations of defects most likely to occur in a given malformation. Of note is the fact that all overlooked VSDs were muscular defects, and more than 50% of the anterior muscular defects were overlooked at the original operation.

Table X. Major associated lesions: age versus mortality Age (mo)

No. of patients

<;3 >3 <;6 >6 <;12 >12 <;24 >24

2 2 2 I 2

Totals

9

Hospital deaths No.

~

T T %

70%CL

75

37%-97%

20

3%-53%

44

24%-66%

} 3*

r}

1*

4

*The p value for difference (Fisher) = 0.17.

Residual shunts should be checked for by obtaining saturation values in the right atrium, pulmonary artery, and radial artery after coming off bypass. Reinstitution of cardiopulmonary bypass and re-exploration of the septum are indicated when important residual shunting is detected. Methods for detection of residual defects in the ventricular septum at the time of repair are currently rather unsatisfactory and better methods are required. Mortality rate. Previously published series have indicated a relatively high hospital mortality rate in patients with multiple VSDs. H. 14-16 We believe that the low mortality rate in the present series in the subsets with VSDs only and TF relates in part to the generally increased safety of repair of these defects in recent years 7, H. 13 and in part to an aggressive search for and re-repair of important residual shunts. The increased mortality rate among patients with

492

The Journal of Thoracic and Cardiovascular Surgery

Kirklin et al.

Table XI. Final known status of interventricular septum Q,,/Qs (n = 20)

Time of last evaluation after final operation

1.0 (closed)

In-hospital * ;=:One year postop. t

3 5

6 3

Totals

8

9

I

>1.0 os 1.5 (small)

I

>1.5 os 2.0 (moderate)

I

>2.0 (large)

Total

3 0

0 0

12:j:

3

0

20§

8

• Determined by oxygen sampling from pulmonary artery. right atrium. and arterial catheters.

t Determined at cardiac catheterization.

*Four patients had both early and late shunt data, and only values from late study are included here. § Nine patients had no available shunt information after the final operation.

Table XII. Probability of Qp/Q s each operative attempt

Category At least I operation At least 2 operations At least 3 operations

:S

2.0 after

Q,,/Qs os 2.0 after that series

No. of patients

No.

I % of total I

70% CL

21* 7 2

14 5 2

67 71 100

53%-78% 45%-90% 39%-100%

• No. of patients in whom Q"/Qs after that series of operations was known. The total exceeds that in Table XI by I because one patient had no QI'/Qs measured after final operation but had an intact septum at autopsy.

more complex malformations and the significant association between the presence of major associated anomalies and hospital mortality rate clearly indicate these subsets to be at particularly high risk for repair. Although there is a suggestion that very young age may increase the risk of operation, the numbers are too small to allow a definitive analysis of this matter. We were unable to demonstrate an effect of ventriculotomy site(s) on hospital mortality rate, but the numbers for left ventriculotomy (n = 6) or combined left and right ventriculotomies at the same operation (n = 4) are too small for firm conclusions. No information is available from our series on potential longterm effects of left ventriculotomies. No incremental risk could be demonstrated with multiple operations or early versus late reoperation (Tables VI and VII) for residual VSDs. Residual shunts. Our early and late studies show that about 45% of patients have small residual shunts and about 15% of patients have moderate residual shunts, even after multiple operations. On the other hand, no patient was left with a final Qp/Qs > 2.0. Thus, if a large residual shunt is present, additional attempts at VSD closure are worthwhile, and there is about a 70% chance of attaining a Qp/Qs :S 2.0 after each operation. The decrease in shunt size from hospital to late study in three patients indicates possible

patch leakage for 24 to 48 hours postoperatively or, perhaps, closure of some small residual defects with time.

Comments On the basis of this analysis, our current management protocol for patients with possible multiple VSDs includes the following: I. Preoperative evaluation includes complete cardiac catheterization with special views to profile accurately all portions of the ventricular septum. 2. Timing of the corrective operation is generally determined by the underlying malformation and patient condition, and not by the presence of multiple VSDs. An exception may be numerous "Swiss cheese" muscular defects, in which initial palliative procedures such as pulmonary banding should be considered in the very young infant. Major associated cardiac lesions and complex malformations significantly increase the risk of operation, and such cases are managed on an individual basis. 3. At operation, the surgical approach is carefully planned for optimal exposure of all known defects. Additional undiagnosed defects should be systematically searched for. Multiple ventriculotomies occasionally may be necessary for accurate identification and closure of all defects. 4. Saturation determinations are made intraoperatively and postoperatively to search for residual shunts in all patients. Large residual shunts (Qp/Qs > 2.0) detected in the operating room are managed by reinstitution of cardiopulmonary bypass and re-examination of the ventricular septum. If detected postoperatively, repeat cardiac catheterization and early reoperation are indicated. 5. Patients are routinely recatheterized about I year after operation to document the hemodynamic result; reoperation for patients with residual shunts in excess of 2: 1 is indicated, even if they are asymptomatic.

Volume 80 Number 4 October. 1980

2

3

4

5

6

7

9

10

II

REFERENCES Taylor lFN, Chrispin AR: Interventricular septal defect shown by left ventricular cine-angiocardiography. Br Heart 1 33: 285- 289, 1971 Young D, Rubinstein B, Schwedel JB: The roentgenographic spectrum in interventricular septal defect. Am 1 Cardiol 5:208-222, 1960 Castaneda AR, Norwood WI: Closure of ventricular septal defects in infancy. Deep hypothermic circulatory arrest technic. Modern Technics in Surgery: Cardiac/ Thorac Surg 1: 1-10, 1979 Kay lH, Anderson RM, Tolentino P, Dykstra P, Shapiro Ml, Meihaus JE, Magidson 0: The surgical repair of high-pressure ventricular septal defect through the right atrium. Surgery 48:65-75, 1960 Kirklin lW, Harshbarger HG, Donald DE, Edwards JE: Surgical correction of ventricular septal defect. Anatomic and technical considerations. 1 THoRAc CARDIOVASC SURG 33:45-59, 1957 Lincoln C, lamieson S, Joseph M, Shinebourne E, Anderson RH: Transatrial repair of ventricular septal defects with reference to their anatomic classification. 1 THORAC CARDIOV ASC SURG 74: 183-190, 1977 Barratt-Boyes BG, Neutze 1M, Clarkson PM, Shardey GC, Brandt PWT: Repair of ventricular septal defect in the first two years of life using profound circulatory arrest techniques. Ann Surg 184:376-390, 1976 8 Blackstone EH, Kirklin lW, Bradley EL, DuShane lW, Appelbaum A: Optimal age and results in repair of large ventricular septal defects. 1 THoRAc CARDIOVASC SURG 72:661-679, 1976 Cartmill TB, DuShane lW, McGoon DC, Kirklin lW: Results of repair of ventricular septal defect. 1 THoRAc CARDIOVASC SURG 52:487-501, 1966 Cooley DA, Garrett HE, Howard HS: The surgical treatment of ventricular septal defect. An analysis of 300 consecutive surgical cases. Prog Cardiovasc Dis 4:312-323, 1962 Lev M, Fell EH, Arcilla R, Weinberg MH: Surgical injury to the conduction system in ventricular septal defect. Am 1 Cardiol 14:464-476, 1964

Surgical management of multiple VSDs

493

12 Lillehei CW, Anderson RC, Eliot RS, Wang Y, Ferlic RM: Pre- and postoperative cardiac catheterization in 200 patients undergoing closure of ventricular septal defects. Surgery 63:69-76, 1968 13 Rein lG, Freed MD, Norwood WI, Castaneda AR: Early and late results of closure of ventricular septal defect in infancy. Ann Thorac Surg 24:19-27, 1977 14 Breckenridge 1M, Stark 1, Waterston 01, Bonham Carter RE: Multiple ventricular septal defects. Ann Thorac Surg 13: I 28-136, 1972 15 Fox KM, Patel RG, Graham GR, Taylor lFN, Stark 1, de Leval MR, Macartney Fl: Multiple and single ventricular septal defect. A clinical and haemodynamic comparison. Br Heart 1 40:141-146, 1978 16 Friedman WF, Mehrizi A, Pusch AL: Multiple muscular ventricular septal defects. Circulation 32:35-42, 1965 17 Werrink AG, Oppenheimer-Dekker A, Monlaert Al: Muscular ventricular septal defects. A reappraisal of the anatomy. Am 1 Cardiol 43:259-264, 1979 18 Shumacker HB: The repair of defects of the muscular portion of the ventricular septum. Surg Gynecol Obstet 136: 103-104, 1973 19 Stewart RW, Kirklin rw, Pacifico AD, Blackstone EH, Bargeron LH: Repair of double-outlet right ventricle. An analysis of 62 cases. 1 THoRAc CARDIOVASC SURG 78:502-514, 1979 20 Fellows KE, Keane lF, Freed MD: Angled views in cineangiocardiography of congenital heart disease. Circulation 56:485-490, 1977 21 Aaron BL, Lower RR: Muscular ventricular septal defect repair made easy. Ann Thorac Surg 19:568-570, 1975 22 Singh AK, de Leval MR, Stark 1: Left ventriculotomy for closure of muscular ventricular septal defects. Treatment of choice. Ann Surg 186:577-580, 1977 23 Waldhausen lA, Herendeen T, Taybi H: Left ventriculotomy. Effects on ventricular function and coronary vessels. Surgery 56:868-873, 1964 24 Zavanella C, Matsuda H, lara F, Subramanian S: Left ventricular approach to multiple ventricular septal defects. Ann Thorac Surg 24:537-543, 1977