CLINICAL PROGRESS
The Natural History of Ventricular Septal
Defect in Patients Surviving Infancy
By DANIEL K. BLOOMFIELD, M.D.
THIS WORK is an effort to define pre-
cisely the natural history of uncompli-
cated ventricular septal defect by approaching
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
the problem from several vantage points.
Despite the medical world's long familiarity
with this malformation, manifest in numerous
studies and reviews of its natural history,
many problems remain unanswered concern-
ing its natural course. What is the long-term
outlook for the patient with the small or mod-
erate-sized defect? What is the lifetime risk of
infective endocarditis? What is the origin of
the pulmonary hypertension in ventricular
septal defect and its relation to pulmonary
vascular changes over the years? What is the
importance, incidence, or even existence of
spontaneous closure of ventricular septal de-
fect? It is unfortunate, academically and prac-
tically, that just as we have entered an era
when information of this nature might be se-
cured, the disease may all but vanish in the
wake of modern surgical treatment. Indeed,
with the surgical risk of repair falling below
the 5-per cent level, we may ask ourselves,
From the National Heart Hospital, London, Eng-
land, and the Department of Medicine, WVestern
Reserve University, Cleveland, Ohio.
* This wvork was planned as a joint project by the
author and Dr. Paul Wood. The tragic death of Dr.
Wood in 1962 occurred before its completion. His
colleagues at the Institute of Cardiology, London,
suggested that Dr. Wood's name should not appear as
an author, since this paper was written after his death.
Nevertheless, aside from the privilege of including the
239 cases of ventricular septal defect personally
studied by Dr. Wood, the author acknowledges his
great debt to the vitality, inspiration, and genius of
Paul Wood, and with deep gratitude dedicates this
work to his memory.
914
Why bother with the natural history? If there
is a proved case of ventricular septal defect it
should be closed. Cooley wrote in 1959, "with-
in the next few years the presence of an un-
complicated ventricular septal defect will be
accepted as sufficient indication to recommend
surgical repair." ` It is well to approach such a
sweeping statement with caution even in the
low-risk groups, for our experience in not com-
pletely understanding ventricular septal de-
fect more than 80 years after Roger's descrip-
tion 2 suggests that medicine will be at least a
generation older before the natural history of
the same disease, complicated by ventricu-
lotomy and Teflon patch, is known.
Methods
This study of ventricular septal defect is based
on four methods of approach: 1. Clinical. Four
hundred twenty-four caseis have been analyzed, of
wshich 185 cases are the total experience of the
National Heart Hospital * from 1948 to 1960 (ex-
cluding cases of Dr. Wood), and 239 cases were
personally studied by Dr. XVood. Data are pre-
sented as representing the total case material
except where specifically designated as the Wood
series or the National Heart Hospital series. Case
material seen has usually been limited to age 3 an-d
beyond, and has not encompassed ventriculam
septal defect in infancy. 2. Hemodynamic. The
defect wvas proved by cardiac catheterization in
288 cases. 3. Follow-up. A subgroup comprising
the first 215 patieints with proved ventricular septal
* I am deeply indebted to the physicians of the
National Heart Hospital-D. E. Bedford, WV. Evans,
C. Hayward, WV. Brigden, A. Leatham, and L. Mc-
Donald-who graciously allowed me to correlate their
patient data with Dr. Paul XVood's material and ac-
tively helped in their collection. Dr. Wood reported
his material in part at the Rachford Lecture, Cincin-
nati, Ohio, October 1961.
Circulation, Volume XXIX, June 1964
 VENTRICULAR SEPTAL DEFECT
defect by catheterization was recalled as a short-
term follow-up study averaging 9.4 years (range
8 to 12 years excluding deaths). In addition, a
long-term follow-up study that averaged 24 years
(range 21 to 26 years, excluding deaths) was
made from 48 patients consecutively selected on
clinical grounds as having ventricular septal defect
when first observed between the years 1934 and
1939. These patients are separate from the clinical
material described above.t 4. Morbid anatomy.
A study of autopsy data noting defect size, cause
of death, incidence of bacterial endocarditis, and
sponitaneous closure was made from a review of
the literature and from a new series of cases com-
piled from the pathology laboratories of four gen-
eral hospitals..+
Ventricular septal defect was studied as a single,
independent lesion. Complicating cardiac and
great vessel lesions a.s.sociated with the defect that
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
might be expected to affect the natural history of
the basic lesion were excluded as far as possible.
The 424 cases of the study include, aside from the
288 cases proved by cardiac catheterization, 136
other cases without pulmonary hypertension hav-
ing adequate clinical criteria. The clinical criteria
for selection of noncatheterized cases are now well
established3 and include the typical harsh pan-
systolic murmur and thrill at the lower left sternal
border, a hyperkinetic left ventricular impulse,
and a mitral rapid inflow mid-diastolic murmur
best heard in expiration. Electrocardiograms often
show prominent Q and tall peaked T waves in the
left precordial leads, and x-rays show enlargement
of the left ventricle, pulmonary artery, and left
atrium, with pulmonary plethora.
Cases excluded from the study were those with
prominent aortic insufficiency, transposition of
the great vessels, atrial septal defect when the
clinical picture of atrial septal defect and the
shunt at atrial level were dominant, single ventri-
cle, and stenosis of the pulmonary outflow tract.
Since patients in the last-mentioned category rep-
resent a special problem in defining the differences
between true congenital obstruction, acquired ob-
struction, and instrumental artifact, the following
standards of selection were made. Patients with
systolic pressure gradients between main pulmo-
t These cases were selected from the records of
D. E. Bedford, who kindly gave permission for their
use, and M. Campbell, retired physician of the Na-
tional Heart Hospital, whose patients are now under
the care of A. Leatham. Dr. Leatham kindly gave the
necessary permission for their inclusion in this study.
+ University and Cleveland Metropolitan General
Hospitals, Cleveland, Ohio, and Massachusetts Gen-
eral and Beth Israel Hospitals, Boston, Massachusetts.
(See table 12, first note.)
Circulation, Volume XXIX, June 1964
915
nary artery and right ventricle greater than 50
mm. Hg were excluded. Included were patients
with a right ventricle-main pulmonary artery sys-
tolic gradient of less than 50 mm. Hg whose pul-
monary blood flow was at least twice systemic
(three cases) and cases in which the gradient
was less than 20 mm. Hg and pulmonary blood
flow exceeded systemic by 50 per cent or more
(three cases).
Other problems of selection included (1) the
separation of left ventricular-right atrial shunts
(a form of ventricular septal defect) from the
double-lesion ventricular and atrial septal defect;
(2) the presence or absence of hemodynamically
significant patent ductus arteriosus; and (3) the
establishment of ventricular septal defect as the
only anomaly in the Eisenmenger group of pa-
tients. Patients with the Eisenmenger complex
were selected oin the basis of Wood's exacting cri-
teria: 4 pulmonary hypertension at systemic level,
pulmonary vas;cular resistance exceeding 10 units;
a reversed or bidirectional shunt at ventricular
level with pulmonary-to-systemic flow ratio of less
than 1.5; oxygen saturation of blood from pulmo-
nary artery less than from aorta; and a ventricular
septal defect of 1.5 to 3.0 cm. in diameter at
autopsy.
Cardiac catheterization data represent 298 stud-
ies in 288 patients over a period of 12 years.
A saline manometer was used to record mean
pressures prior to 1952 (17 cases), after which
an electromanometer was used. Patients were usu-
ally mildly sedated except for a few cases of
younger children, who were anesthetized. Oxygen
content was measured by the Haldane technic.
Catheter reports and tracings, were reviewed
and in somie cases recalculated so that figures
could be compared to a common baseline. Stand-
ard assumptions were made when catheter data
were incomplete (in detail, not in diagnostic val-
ue). These assumptions, based on mean values of
the same data from 21 or more patients in whom
the specific figure was measured or recorded in
the laboratories of the National Heart Hospital,
include the following four. 1. Oxygen consumption
equaled 1.28 times the basal metabolic rate
derived from tables.5 2. Pulmonary vein oxygen
saturation for computing arteriovenous difference
was assumed to be 95 per cent. 3. Systemic ar-
teriovenous difference was calculated from brachi-
al artery saturation and the mixed venous sample
from !the lateral wall of mid-right atrium. When
left ventricle-right atrium shunt was diagnosed
(three cases), systemic flow was calculated on
the basis of brachial artery-superior vena cava
difference. 4. Pulmonary resistance, expressed as
units, was calculated as the quotient between the
observed difference between mean pulmonary
artery and pulmonary arterial wedge pressures in
 916 BLOOMFIELD

mm. Hg and the pulmoniary flow expressed in
liiters per minute. The upper limit of normal
resistance is about 2 units.
In the follow-up studies all patients were traced
through their former residence and through their
referring physician. When this failed the name
of the patient was turnied over to a volunteer
aglency * specializing in the location of missing
* I am extremely grateful to Lt. Colonel L. Clayson
Thomas, Chief, The International Investigation De-
partment of The Salvation Army, London, and his
staff through whose generous assistance were located
plersons. Siince the paramount question in loiig-
term follow-up ventricular septal defect is patient
survival, the name of each patient not located was
searched in the Register of Deaths, General Regis-
ter Office, Somerset House, Londoin, Englaind.
This office records all deaths in the United Kinig-
doin and the absence from the files of the names
of all patients not located in the series to be re-
ported below is firm evidence that the missinig
several of the living persons in these studies, and
who supervised the difficult search through the Regis-
ter of Deaths, Somerset House.

Table 1
Ventricular Septal Defect-Classification and Hernodyniamics
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

Uncomplicated by Complicated by
elevated pulmonary vascular resistance elevated pulmonary vascular resistance
Mtild Moderate Severe Pulmonary hypertensive Eisenmenger complex
Group no. 1 3 4 5
No. cases 175 98 18 63 70
Per cent of total 41.2 23.2 4.2 14.9 16.5
Age
Average 16.4 13.8 17.2 10.3 20.8
Median 14 10 11 6 22
Range 3 -71 1 -44 3 -35 1 -25 5 -44
Pf: Sf flow ratio * 1.1 - 1.7 1.8 - 2.9 >3.0 1.5 - 3.5 0.5 - 1.5
RV systolic
pressure (mm.Hg)
Average 21 24 35 82 Systemic
Range 12-30 15-40 20-70 60 - systemic
Pulmonary artery mean pressure (mm.Hg)
Average 13 18 38 56 Systemic
Range 5 -20 5 -30 12 -55 35 -systemic Systemic
Pulmonary vascular
resistance, units
Average 1.25 1.1 2.0 6.2 20
Range 0.4 -2.2 0.4 - 2.5 0.8 - 4.0 3-10 10 - 40+
Hyperkinetic pulmonary None None to Moderate Severe None
hypertension mild to severe
Size of defect (cm.)
Average 0.4 1.0 2.0 2.2 2.2
Range 0.1-0.6 0.7-1.2 1.5-2.0+ 1.5 - 3.0 1.5 - 3.0
*
Pulmonary-to-systemic flow ratio.
Ctrcuiation, Volume XXIX, June 1964
 VENTRICULAR SEPTAL DEFECT
32[
28
(I)
z o
-24
U A
z
20
16i
g 12
0
U)
>~8
*
* . .I..
z 4
0 .-:a
A * k -*
i :
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
a-
2 3
PULMONARY TO SYSTEMIC FLOW RATIO (Pf:Sf)
Relationship between pulmonary vascular resistance and pulmonary-to-systemic
ventricutlar septal defect.
patients have not died within the limits of the
United Kingdom. The deaths of three patients in-
cluded in these studies were uncovered by these
searches. This type of follow-up accentuates
prognosis.
Further details on methodology are described
as they arise and bear upon the results reported
below.
Results
Classification
Ventricular septal defect is ideally classified
on the basis of defect size, amount and direc-
tion of shunt, and resistance to flow in the
lesser circulation. Cardiac catheterization pro-
vides satisfactory indices of these criteria in
most situations, but examples will be cited
when clinical classification gives results closer
to the truth. Careful attention to brachial
pulse, impulses of right and left ventricle and
pulmonary artery, quality of the second sound,
and mitral inflow murmur together with the
interpretation of electrocardiogram and x-ray
is also an accurate method of assessment.
Ventricular septal defect has been classified
as complicated by elevated pulmonary vascu-
lar resistance or uncomplicated. Uncompli-
cated cases are mild, moderate, or severe.
Complicated cases are characterized by hyper-
Circulation, Volume XXIX, June 1964
917
* GROUPS I& 2
GROUP 3
a
GROUP 4
GROUP 5
m
:Ug
0 0
oo
0 0
0 0
0 0 0 0
.0 0
4 5
Figure 1
flow ratios in
kinetic pulmonary hypertension with moder-
ate or large left-to-right shunts, or without
hyperkinetic pulmonary hypertension, pul-
a bad
monary artery pressures at systemic levels,
and various degrees of bidirectional shunt
(Eisenmenger complex) .4 These five groups
are characterized in table 1. A significant fea-
ture of table 1 is that 64.4 per cent of cases fell
into the mild to moderate classification, another
16.5 per cent into group 5 (Eisenmenger com-
plex), and only 19.1 per cent in groups 3 and
4, in which surgical closure has proven value.
Figure 1 is a plot of pulmonary-to-systemic
flow ratios against pulmonary resistance in
units. The separate groups aggregate in a dis-
tinct pattern. Groups 1 and 2 blend imper-
ceptibly with each other and form a unit dis-
tinct from the continuum of points that define
groups 3, 4, and 5. One inference to be made
from figure 1, which is substantiated by data
presented below is that groups 1 and 2 do not
blend into the pulmonary hypertensive groups.
There are few borderline cases. Since re-
sistance varies inversely with flow, groups 3,
4, and 5 naturally form a hyperbolic curve.
The points at issue to be analyzed are what
determines where a given patient lies on this
 918 BLOOMFIELD
Table 2 Table 3
Age Incidence in 424 Cases of Ventricular Septal Comparison of Age Incidence between Venttricular
Defect Septal Defects, Groups 1 and 5, Demonstrating
a Significantly Higher Fraction of Group 5 Found
Age at first visit after Age 30
Group 1-10 1 1-20 21-30 31-40 41 -50 51+ Total
Cases Cases Total
1 65 65 24 14 4 3 175 ulnder 30 over 30 110.
2 43 40 8 5 2 0 98 years years cases
No. % No. %
3 3 6 2 0 0 18
4 36 18 7 2 0 0 63 Group 1 154 88.0 21 12.0 175
5 20 14 20 9 6 1 70 Group 5 54 77.1 16 22.9 70
Total 171 140 65 32 12 4 424 Total 208 84.9 37 15.1 245
X2 4.59
1
curve and what factors, if any, influence P = <0.032
changes of that position. These cannot be re-
solved before other features of natural history patients over 30 with the mildest form of de-
are understood. fect (group 1) is significantly less than that
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

of patients with the Eisenmenger complex. To

Age relate this to survivorship for either group is
The age incidence when patients were first nonsense. Failure to appreciate these facts
observed is presented in table 2. Fractional has led to the confusion surrounding the nat-
distribution of these cases is shown in figure ural history of this defect. For this reason, age
2. However, the age at diagnosis for any dis- incidence also is discussed with, and in the
ease or the age at which a patient first appears light of, data that follow.
in a clinic is poorly related to survivorship. For
example, the age incidence of patients with Sex
osteoarthritic complaints roughly parallels that Cases were distributed evenly between
of myocardial infarction but prognosis for the sexes in all groups at all ages. Elevation of
two is clearly unrelated. Similarly, it can be pulmonary vascular resistance was unrelated
shown (table 3) that the age incidence of to the sex of the patient (table 4).
Growth and Development
601 The height and weight of 70 catheterized
cases from the National Heart Hospital series
i'50 who were below age 21 were compared to
standard reference tables.0 The two cases in
,X40
0
group 3 in which data were available were
combined with 19 from group 4. Each case
30
.03 was plotted as being under the 10th percentile
0
for age, between the 10th and 50th percentile,
~ 20 or over the 50th percentile. Results are pre-
a
-0
I0
Table 4
Sex in Ventriculat Septal Defect
_.. _-
Male Feinale
10 20 30 40 50 60
Age Groups by Decades Uncomplicated 154 137
(groups 1 to 3)
Figure 2 With elevated pulmonary 64 69
vascular resistance
Fractional age distribution by group for defect. The
(groups 4, 5)
late peak of group 5 is statistically significant; the
late peak of group 3 is not. Total 218 206

Circulation, Voluwme XXIX, June 1964

 VENTRICULAR SEPTAL DEFECT 919
PERCENTILE
100
a <10
10-50 80
: >50
U)
Uf)
< 40-
U) _ 40
1
20 520

N9.CASEES 5 3 7 7 17 4 10 10 5 0
0% GROUP 2 3&4 5 NO. CASES 2 7 6
VSD GROUP 2 3&4 5
Figure 3 Figure 4
Relationship between growth (height) and defect Relationship between growth (weight) and defect
group in patients under 21 years of age. Growth is group in patients under 21 years of age. Symbols are
significantly retarded in patients with hyperkinetic defined in figure 3.
pulmonary hypertension (groups 3 and 4).
Eight- to 12-Year Follow-up Study on the
sented in figures 3 and 4. In all groups except First 25 Patients Catheterized at National
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

the first (maladie de Roger) there is a tend- Heart Hospital with Simple Ventricular Septal
ency to smaller stature. This is most marked Defect [Eisenmenger Complex (Group 5)
Excluded]
in groups 3 and 4 in which half the children To achieve the longest possible follow-up
were below the 10th percentile in both height
and weight. Even among group-i patients, of definite cases, the first 25 patients with
five of 15 were below the 10th percentile in catheter-proven lesions (excluding Eisenmen-
height, leading to speculation that these pa- ger complex) were selected for study (table
tients may have had larger defects earlier in 6). These catheterizations were done between
life. Eight older patients with Eisenmenger 1948 and 1952. Eighty-four per cent (21 pa-
complex (group 5), ages 21 to 44, were small tients) of the 25 were located; 100 per cent of
in stature. None of these patients, all female, the pulmonary hypertensive groups. The fol-
was taller than 63.5 inches (range 58 to 63.5 low-up study included a hospital visit with
inches). clinical examination by the author, electro-
The small stature may be related to a lower cardiogram, and x-ray in 12 cases; home visit
cardiac output. Cardiac indices for each group by the author, one case; report from private
are tabulated in table 5. While average ages physician, four cases; autopsy report, four
are not strictly comparable, the mean cardiac cases.
index for groups 2 to 5 is lower than that
for group 1. These differences are significant Group 1
at the 5-per cent level or less except for group All patients (average follow-up 9.2 years)
4. Group 4 patients, however, are those with remained stable. No changes occurred in
significantly smaller stature than group 1 and symptomatology, physical examination, elec-
age-for-age cardiac output is lower. trocardiogram, or chest x-ray.
Table 5
Cardiac Index in Ventricular Septal Defect (National Heart Hospital Series)
Mean
cardiac Probability
Average Median index, of difference
Group No. cases age age L. / min. I M.2 from group 1
1 16 12.4 9.5 5.32 + 1.9
2 34 16.1 14.0 3.80 1.0 <0.005
3 6 18.0 20.5 3.28 + 0.5 <0.001
4 21 10.2 7.0 4.41 + 1.0 <0.10
5 14 20.7 21.0 3.80 + 1.5 <0.01

Circulation, Volucme XXIX, June 196(4

 920 BLOOMFIELD
Grouip 2
r,a The eight cases of group 2 (average follow-
up 10.1 years) separated into three subgroups.
One patient improved:
Case Report. An asymptomatic young woman
doI~> v of 28 was referred to National Heart Hospital in
1950 because of persistent awareness of her heart
8e murmur and active precordium. Her physical
examination was consistent with a mild-to-mod-
erate defect and cardiac catheterization confirmed
this, showing a pulmonary-to-systemic flow ratio
of 2.0. The patient was lost to follow-up, but
was located in 1960. When visited at home, she
admitted that she had not thought it important
0 to be seen again, since her murmur (which she
had felt as a thrill) and active precordium had
diminished remarkably in the past decade. Physi-
cal examination showed only a grade-IlI pan-
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

systolic murmur in the Roger area accompanied

a by a faint thrill. The precordium and heart size
0 1 were normal. This case, therefore, represents a
1 .1~ patient in group 2 entering group 1 between her
twenty-eighth and thirty-eighth years.
Four of the eight cases located were un-
_ ~-
changed. The three remaining cases in group 2
were worse but not because of increased pul-
* **
C
z.m > C r1 cO 0
Cld
monary vascular disease. One of these, re-
maining asymptomatic, developed increased
W
aortic regurgitation between ages 11 and 21.
0+i The patient had been selected initially because
0
et
CDC
a
a
01D .
the aortic regurgitation was minimal at age 11.
. C
N A second patient noted more dyspnea on ex-
a ertion between ages 41 anid 49. Although her
t
c o W CX x-rays were unchanged, her electrocardiogram
showed increased right ventricular hypertro-
, ,
phy. Cardiac catheterization showed a remark-
CO (1 = > able increase in pressure gradient at infuindib-
ular level (case DC-2, table 7). The small
right ventricle-pulmonary artery gradient of
0 l CO' e 15 mm. Hg, thought to be valvular at the time
I OE
~I -
of first catheterization, increased to a 74-mm.
_~ oO 1X Jc
Jig infundibular gradient between ages 41 and
mmC7 C0) 1. zI,c'^ 49. The left-to-right shunt diminished slightly,
0m ibk- "
tudZ _ D and pulmonary vascular resistance remained
.o1, i~A C>11
the same. This patient demonstrates that even
0 in mild instances of pulmonary stenosis with
ventricular septal defect the pressure gradient
across the pulmonic valve may increase late in
CP life. The third case showing an unfavorable
ci c) io 1
. -
change over an ll-year period (case DC-3,
table 7) was 15 years old when first observed.
Her first catheterization showed a pulmonary-
(Gir u//io Volume XXIX. June 17964
 VENTRICULAR SEPTAL DEFECT 921

d .H C)
Ca
GO
Ca
C)
C)
co cn Ca
Ca
(L) crQ a)
ct Ca
t01 0
Ui:: 04
CN
.ca A
C)
QO
co ca
.tc-Ca
'-GO a)
1-
'-G
W
r_, ot
o uz .

a) 0 cm
Ca
CaG can
"'a)
4.i C) C)
Ca "GO a1) C0
0 bfo GOG
$_4 Ct o Ca COD
ca
Ca . i
U) ct
- ;Q
-4
Uj
C4GO W
C) W- r. ~ca
04 $_4 a) C) '-G ;"GO
CD a) GO
Ct a)
C cn
U:Ct
~r.(La ,.CJ
"GO
a) 8Z > *14
GGo ---)
GO
S} E o ct 0
C C) -cl 491 . rO
a)
Enl -;
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

CL.
Ca
._ <
xpnls
4- +++ ++
Ct
o.
0 C0 : 0 M -
00 C O00C C
d O -; d t ed o6 c 010
c00
af-ZI~ b~
CO 0 *0
aU)
CO - CO o0
o F

001C C C O0
C cOn4j co
Ca
0110 01 GO
01 010 01 Co
GO 01 C
C) \1 'N
Car
b __ 01 CO CO 01 GO a) C,
C) _0 CQ _i _ 00~C C )
C) C00 - 10
bn
- b~~~
I I
Vd - cc
Io It-o I 01 o o o IoIo
0.
C) S
uea
00 10 ++ 0 m 00
a)
,,
U-COI'MJ 0
oCo
010
-c s
10
01c oo _00
o 00 m 0o
,6
C)
Vdl
0 c01n co
- _ 0000I 100o10 Ca4 * P-4
C) Cc b> 00 CO 1-00 00 1- 0 0 C

C-
OinC_-ci
10 CcOc
n C: t- 1 D 00M 00M
o rl
0o01 10 10 0 0 X 10
m 01ol11 o 1- 4
A) S 0 n co CD 01 0 o Io I 110010O
C"
1- 1- M

CD ce) b b a) ct
ao
GO
C) vi 00m'
1 1 CO CO O t- O7 >
oG GO; o

GO ._os I' o' o ct

MiVe O CO o 0
00 -0000
0'10o0r.'- I 00 0o 04 a))Ca
0
Ca
~,1 In
001m 0 00 7
1-
01 CO CD 10I
0 00 j>- 0000 Cdd X
GO va- > >
Ca a Q;>
GO CD cn " 1n
ICO 0'
M
00>1 10' I1-40_ O- c Ca Ca<
C) xas - 01 C 00 0
oo 1001000
C5) oO= Ca"
GO a)a*"=
GO a2 _ t- Cq - 0
c- oo o- n -onn
-
0
C) 01 Cl 10 1- 00 t "e 0 GO.E
Q CU CU U ) Ca v

a >* o)Wj
C)
00 dno.iD CO 4-ca

Circulat;on, Volume XXIX, June 1964

 ( 2)
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
Figure 5A andl B
X-rays of a grtotp)-3 prtietw (case SC-1, tril)le 5)it/iO
puilmrinaryJ lIypertentusio (htyp)erkincttic) aid near -normall p)ulmtlon(ary tasctlr resistance(at ages 24
rind 28. A, left. X-ray, agec 24. 13, right. X-rayl, age 28, fnange fottl age( 24.
to-systemic flow ratio of 1.8, less thaln antici-
pated fromn physical examiniation of ler active
precordiuim. Eleven years later she was again
observed to have a hyperkinietic precordium
stuggestive of a large shunt. Chest x-ray xvas
unchanged, buit an electrocardiogram shoxved
evideniee of inereased left ventricuilar diastolic
overload. The pulmonary-to-systemic flow ra-
tio at this time was 2.9 anid her calculated
ptulmonary vascular resistance had halved.
Since her brachial artery satuiration was the
same at both catheterizationis (94 per cent
versuis 95 per cent), the results cannot be
explaine(l on the basis of anoxia at the time of
ler first catheterization. The catheterization
data from this case suggest progressi.on of a
moderate defect to the severe form, uncom-
plicated by pulmonary hypertension. The aui-
thor helieves, hOwexver, that this patient al-
ready exhibited the severe, uncomplicated
form of defect (grouip 3) when she was first
seen at age 15. The left-to-riglht shuint calcu-
lated from catheterization at that time was
inconsistent with the clinical description of
her markedly hyperkinetic heart.
Gr.o)up 3
Tim three patients from this grotup exhibited
varie(l hi it g(en erallv unfavorable courses.
92 BOOL)IFIEIL)
lwhhd large pulmonrrs flowc, moderate
Serial catleter izationi dlata are available for
two of these in table 7 (cases DC-4 and D- C-
5). All lhad very large shutnts anid low Pull-
mnonary resistance.
Case Report. Onle mani (ease SC-1, table 8)
clied of his disease at age 35. Dr. Wood's patient,
lhe,had the most ma kedly hyperkinietic left
venitr-icular impullses Dr. WVood had ever seen
in a x enttricular septal (lefect xx itholut associaxted
aortic insuifficienlcv. Ther e xwas a veiv large ill-
criea.se inl oxvgel) satlurationi b)et\x eeii right atrium
and right venitrlicle. The ptulmoniar-v-to-svstemic
flox ratio xxas 4.2. Typical for this group, cardiaic
index xwas loxw, 2.6 L. mm. MY. His chiest x-rays
at ages 24 and 28 are slhowni in figure 15. I is
electrocardiogram at age 28 (fig. .SC ) shoxved i-
ventriicuilar hvper-tr-ophxvx with per sistence of (leep
Q and tall R wxvaes in left precordial leads. His
death and IneerIopsv aIt anilother hospital sever-al
veairs later xvere no.t well docuimenited at the time,
but after a later reviewx xxith his xidox, priivate
phvsicianii, annc proseetor, of events precedinig
death, it seemed cleai- that left venitrlic-ulla failur.e
wvas a primiary catse. Hle had moderate dxyspnea
otn exertion and. orthopnea the laist 6 molths of hiis
life, bhit he was nxeer cyanotie. He had niolt sought
mnediecal car e for the 6 v ears prior to hiis dleathi
and had called hi.s phvsician onlx 3 dlays before
the end, because. of couighi, fever., and dvspnea.
He had nexver been dligitalized. At necropsx
the heart wxeighed 940 Gm., the lhugest hlearit
in- a death from simple ventricular septail cle-
C11O, 11oliimt \X1I. J1/un 1964
 VENTRICULAR SEPTAL DEFECT
I v,
V4
mI
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
V6
Figure 5C
Electrocardiogram, (case SC-1) aged 28, showing
biverntricular hypertrophy.
fect yet reported. There was general chamber
enlargement, most marked in the left ventricle.
The mitral and tricuspid valves admitted "four
fingers," the pulmonary valve "two fingers." There
was a "large" high defect with apparent overriding
of the aorta. The lungs were edematous with
widespread bronchopneumonia. Histologic sec-
tions of the lung were not available at the time
of inquiry (1 year after death) and no observa-
tions were made concerning the pulmonary vas-
culature. In summary, this case represents a
colossal left-to-right shunt at ventricular level
functioning for 35 years without clinical evidence
of progressive pulmonary vascular disease that
would have led to shunt reversial.
Case Report. The second case in group 3, al-
though complicated by mild pulmonary valvular
stenosis, was a woman followed for 8 years be-
tween the ages of 50 and 58 (case DC-5, table 7).
In 1960, because of continued but not progressive
symptoms, she demanded and underwent open-
heart surgery for correction of her defect. At
surgery a 2-cm. defect and mild pulmonary
valvular stenosis were corrected. Recovery was
delayed by a transient hemiplegia, and at a post-
surgical examination it was apparent that her
Circulation, Volume XXIX, June 1964
923
shunt had at least partially reopened. Neverthe-
less, she was symptomatically improved within 2
months of surgery. Although her pulmonary vas-
culature was in part protected by the mild pul-
monary stenosis, her shunt with a calculated pul-
monary-to-systemic flow ratio of 5.0 that had
operated for 58 years had not produced irreversi-
ble pulmonary vascular disease.
Case Report. The third case in this group (case
DC-4, table 7) is another example of the difficulty
in evaluating pulmonary changes without the
routine administration of oxygen in such in-
stances. This patient had no symptomatic change
over the 10-year period (ages 21 to 31) he was
observed. He had always been breathless on heavy
exertion. Chest x-ray was also unchanged. Electro-
cardiogram showed slight axis shift to the right
and higher precordial voltages consistent with
increased right ventricular strain. The two sets of
catheter figures 10 years apart demonstrated an
increase in pulmonary vascular resistance from
1.1 to 4.7 units. Resistance fell to 2.3 units with
oxygen administration, but the mean pulmonary
artery pressure remained 7 to 11 mm. Hg above
the initial figure of 10 years before. This is an ex-
ample of the gradual, but definite, progressive
increase in pulmonary vascular resistance in a
ventricular septal defect with a large shunt and
moderate pulmonary hyperkinetic hypertension.
The change occurred after 21 years of high flow
and low resistance.
Group 4
Although four of the five patients had died
at the time of follow-up, only one died as a
direct result of increased pulmonary vascu-
lar disease and right-heart failure. A second
(case DC-8, table 7) died shortly after surgical
closure of the defect. The third patient died
of bacterial endocarditis at age 8, and the
fourth death occurred in another boy (cf.
group 2 follow-up, above) who was included
in this series only because the aortic insuffi-
ciency associated with his defect was thought
to be minor when he was first seen at age 13.
Clinical and electrocardiographic right ven-
tricular predominance was observed at that
time. Letters from the father of this patient,
not a physician, 15 months after his son's
death at age 21 described the final course of
the disease: "A. was just slightly blue around
the mouth before he died. . . . He was always
short of breath and could walk only about 20
yards. He could not lie flat; we had a board
at the head of his bed with about six pillows.
 1
924 BLOOMFIELD
Regarding chest pains, he was never free of
thelm.... When he was in bed it always shook !1
11
-1;
Q. .r,
4
with the movements of his body." There was _. .z
1-
r,.,1
;
no necropsy, but the above description and a 1,1
P..
11 Wv
letter from the private physician imply that
aortic insufficiency played a major role in the
patient's demise.
.'Ii
The only patient in group 4 that "improved" <i)
x Z)
~ >
during this study (case DC-9, table 7) most r Wt
likely represents the type of artifact that
anoxia can produce in catheterizations. The I~ ;c
CC' c) c
findings, at her first catheterization, of equal C C
pulmonary and systemic pressures with bidi-
rectional shunt placed the patient in the Eisen-
menger complex group. This was contrary to
bedside clinical appraisal, which suggested a Circidwion, Volume ..X J e1
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

moderate left-to-right shunt with pulmonary

hypertension. Seven years later she was clin- c

ically the same, although x-ray showed in-

creased heart size. A second catheterization
placed her in group 4. ~~~~~ A; -
~~ ~ m
b4
Twenty-one to 26-Year Follow-up Study of
Patients with Possible Ventricular Septal Defect
A further clue to survivorship and long-
term prognosis is offered by follow-up obser- 0: p z O

vations of a group of 48 patients selected 11

C] C) 0 '
from 4,200 consecutive records of out-patients
seen in the clinics of two consultant physicians
of the National Heart Hospital (Drs. Bedford
and Campbell) between the years 1934 and 1K
(>

1939. The average period of follow-up was 24

years. Diagnosis in 1939 was necessarily clin- r. -

ical. These patients were selected as probably C t- l:D -

having a defect on the basis of clinical descrip- !
tion, 3-lead electrocardiogram, and fluoro-
I
scopic reports. Tracings of cardiac outlines C'I

from fluoroscopic screening were available in

20 per cent of the cases.
Patients were classified into three severity
groups on the basis of evidence in the old X a)
charts: mild (severity group 1, 29 cases),
asymptomatic but having the typical harsh, 1 :
pansystolic, left lower sternal border murmur t ~~~~~~
z X
without clinical evidence of a shunt; moderate
(severity group 2, 17 cases), in which hyper- ~ ~ ~C'U
, ._ _
kinesis suggesting a shunt was present by 0S 1
physical examination or fluoroscopy, yet with- i
(z';dauonv

out symptoms; and severe (group 3, 2 cases), x3 1

in which the signs of the defect were present i;10
and the patients were symptomatic.
 VENTRICULAR SEPTAL DEFECT
Each case was also classified as to the proba-
bility of the original diagnosis being correct.
This was considered excellent (group A) in 12
cases, good (group B) in 19 cases, and fair
(group C) in 17 cases. The "excellent"' cases
fulfilled all the classical criteria for this dis-
ease, while those labeled "fair" included pa-
tients where assurance of diagnosis was lim-
ited by some of the following details: age un-
der 3 years; absence of thrill noted; possibility
of pulmonary stenosis; diagnosis of "systolic
murmur" rather than ventricular septal defect
or congenital morbus cordis; preceding history
of rheumatic fever; murmur first heard after
age 5; and the presence of signs of pulmonary
hypertension in the two symptomatic patients.
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
Group B cases fell between the other two.
Twenty-nine cases (60 per cent) were lo-
cated. Five of these 29 (17 per cent) were
shown in 1960 to have been selected in error.
Two had pulmonary stenosis proved by cardi-
ac catheterization and another had atrial sep-
tal defect, ostium primum. One was a 9-year-
old boy who presented with aortic stenosis and
insufficiency when seen 25 years later, and the
fifth case had the murmur and findings of mild
mitral insufficiency when seen in 1960. Two of
the five cases diagnosed in error (atrial septal
defect and aortic valve disease) were in prob-
ability group B and the other three were in
group C.
To eliminate the chance that selective fac-
tors, other than a search of the Register of
Deaths, were weighting follow-up data, the
Table 9
Demonstration by Several Clinical Factors of the Nonselective Nature of Patient
Follow-up in the 24-Year Study
1934-1939
Total cases 48
Male 21
Female 27
VSD severity
Group 1 29
Group 2 17
Group 3/4 2 2
Thrill
Present 24
Absent 6 4
Circulati'onz, Volume XXIX, June 1964
925
48 cases were subdivided according to sex,
age, severity and probability groups, and spe-
cific clinical or historic criteria. The per cent
follow-up for each subdivision was roughly
parallel to the fractional recovery of the 48-
patient group as a whole (table 9).
A summary of the results of this series is
presented in tables 10 and 11. Twenty-nine of
the 48 patients (60 per cent) were located
alive or dead. Eighteen were examined by
the author at home or at the National Heart
Hospital, seven were examined by private
physicians or at other hospitals, and four were
reported from nonmedical sources. As indi-
cated previously, a search of the Register of
Deaths was completed for the 21 patients not
located, so that we can say with confidence
that while they are missing from our follow-
up, they have not died within the borders of
the United Kingdom.
Fifteen of the 29 patients (52 per cent) had
some degree of cardiac pathology, whereas
14 (48 per cent) had normal hearts in 1960.
Since five of the 29 patients (17 per cent) are
known to have been selected in error, the
data discussed pertain to the 24 who remain as
possibly having had a ventricular septal defect
24 years before. The diagnosis of defect was
confirmed in 1960 for eight patients, as
follows: two by cardiac catheterization,
one by autopsy, and five by clinical
evaluation. Among 13 who had a thrill
noted, six (46 per cent) were normal in 1960.
Of the 14 patients with a mild defect, 10 (71
Number in this
classification Per cent
located in 1960 follow-up
29 60
13 62
16 59
18 62
9 53
100
16 67
67
 .\2,EOz/R{CoUs;i'w^a>*4_#-AXt)b:eWV
926 BLOOMFIELD

.! .it per cent) were normal and four (29 per cent)
1~~~~~~~~~~~~~~~~~~~~~~~~~
,
1w liad the saine pathology. Of the eight patients
1l
c whio were classified in the severity group 2.
tour (.50 per cent) were normal, one (12 per
cent) had the same degree of defect, two (25
,04
P:
r', 1%
per cent) had a lesser degree of defect, and
Q1
C
.
-> M1 th-e eighth was a patient with autopsy-proved
ventricular septal defect and bacterial endo-
_ . Ogl
1, 0 Clt C5cc
5 Cl
rl carditis. Her death occurred 15 months after
0 I0 N her visit in 1937. The patient originally pre-
h sented with fever, anemia, and tachycardia,
~ .- r] c--
and the diagnosis of subacute bacterial endo-
1 U u0
0 0cc carditis had been suspected at that time, but
she was lost to follow-up until this study was
doine. Her death and autopsy report were un-
0 cc 0 covered in the search of the Register of
t zCl!C
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

C
Deaths.
0 C
Dl > , i , The two patients in group 3 died 4 and 11
0
years after first being observed; both were in
probability group C. One was a girl 3 months
..+ Cl
: t_
,.

5:
of age when first seen. Her autopsy records
Cc were lost in the bombing of London during
.;
_
WYorld WVar II and death at age 4 was recorded
4 Q t? S- io I C)
c Cc'
:r
as duie to "mongolism and tuberculosis of lung
t-4
r and skin." The second was a 30-year-old
woman wlho died 11 years after her visit.
1_

Qf S r5
0
Cl
I as Atitopsy was not performed. Death certificate
y
I rr . ~
recorded "myocardial degeneration and cere-
ry
bral vascular accident." These two cases prob-
S CN ably should not have been selected for this
:r type of follow-up study, because of the diffi-
culty of clinical diagnosis of the symptomatic
._

Cc : . _
'' ~ C ccC f G) ,
cn .
'S ,\U b Cr fi
patient with cardiomegaly in 1937. While not
i ,_
-_

tt;~~~C
helpful with the natural history, they do help
_ .

L0|
3
U to establish the validity of the assumption
. that the search of death records was complete,
and follow-up has not overlooked other deaths.
_ wv ,
! ~~ ~ C Cl
0j I D sF+ i,-f > Cl' _9 . Q _
Ilemoclyianutic Changes
In the mild and moderate groups all hemo-
J M K| O 1lC.C v rrH aS _ . dyniamic changes, as clinically determined,
0 t. . >
fi t z R
were toward amelioration or disappearance of
the defect. Electrocardiograms were un-
changed in tlhe eight patients with tracings
available before and after follow-up. No pa-
tient in these groups developed pulmonary
g 4
_
,'_
J
Li
:1
.
Cl
-
-X Iypertension. This is consistent with other
-
V = = i '
O' Gi W ;)
observations-that smaller defects never de-
v elop ptulmonary vasetl.ar complications.
* + ts ;<> w

C,rV.idattVn.I/olumrne XXIX, Juine 1964

 VENTRICULAR SEPTAL DEFECT
Relation of Age and Grouping to Persistence of Murmurs and Possible Spontaneous
Closure of Ventricular Septal Defect in 24 Patients Followed 24 Years
VSD
severity Age, yr.
group 1934-1939
0-4
5-10
Group 1 11-20
>20
0-4
5-10
Group 2 11-20
>20
3/12
Group 3 30
*
Includes patient dying in 1937.
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
Spontaneous Closure of Ventricular Septal Defect
in a 24-Year Follow-Up
The data presented in table 10 emphasize
the benign prognosis of the mild defect and
suggest that the defect may become smaller
or disappear. Only one patient in the mild or
moderate groups died, and she probably pre-
sented herself in 1937 because of symptomatic
bacterial endocarditis. The possibility of spon-
taneous closure of these defects has intrigued
physicians since the early report of Weber7
and for this reason the information from the
24 patients located who had a possible lesion
was broken down by age group and severity
to gain further insight into the problem (table
11). It will be noted that the disappearance
of these loud murmurs was not limited to the
first decade of life. Six of 10 survivors in
groups 1 and 2, who were older than 10 years
of age when first seen lost their murmurs (or
closed spontaneously) in the next 24 years.
Further data concerning spontaneous closure
is presented below, as is the incidence
of bacterial endocarditis.
A feature of this study, which is the key to
the most important reason for the paucity in
our clinics of older patients with milder de-
fects, concerns the seven patients (average age
40.6 years) located who still had signs of a de-
fect. Only one of the seven was being followed
regularly by a physician. Three knew tlhat they
had a "heart murmur" but did not seek medi-
cal care (one being a patient who was suc-
Circulation, Volume XXIX, June 1964
927
Table 11
Status in 1960 or at death
No. VSD
patients present Normal Dead
6 1 5 0
2 1 1 0
5 1 4 0
1 1 0 0
0 0 0 0
3 1 2 0
1 0 1 0
4 3* 1 1
1 0 1
1 0 1
cessfully treated for subacute bacterial en-
docarditis at another hospital in 1950 but had
not sought follow-up care since 1952). Three
were entirely unaware that a heart murmur
was present. In other words, six of seven pa-
tients with a defect at age 40 might be
missed if a general survey of all clinical ma-
terial available were studied. Furthermore,
these asymptomatic healthy patients were
typically unwilling to go out of their way for
medical examination. Only eight of the 18
patients personally examined agreed to visit
the hospital, although, to be sure, distances
were in some instances forbidding. The re-
maining ones were examined at home or place
of work. When the author visited them, all
were cooperative and apologized for their de-
lays, but their well being as adults had con-
vinced them that they did not need to visit a
doctor. This was true; the doctor, anxious to
complete the study, needed to visit them.
Acceptance of this type of follow-up de-
pends heavily upon the validity of the retro-
spective diagnosis of two and one-half decades
ago. The pitfalls in the correct diagnosis of
defect-like murmurs have been described pre-
viously.8
The differential diagnosis of milder forms of
defect, particularly in younger and certainly in
older age groups, is often difficult. Pulmonary
valve or mild infundibular stenosis, mild aortic
stenosis, mitral insufficiency, acyanotic Fal-
lot's tetralogy, and patent ductus arteriosus
 928 BLOOMFIELD

601, KEY
a)
50
.
* DEATH RELATED to VSD a PHT
E
E 40 . 0 DEATH UNRELATED
6 30 o CAUSE NOT RECORDED
C.)_ . a . * BACTERIAL ENDOCARDITIS
0
* 0
0

CD)
0) 20
W 0 O U
U) O-

-0
0 E:*
0 a
c
_c]
c: 10 C]

0~~~~~~~~~~~
m0 0
(n cloc

Cl
0 .~~~~~~~~~~~~
a~~~~~
CP 'I
cl
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

A I * AI l l I l
1 2 3 4 5 6 7 8 9 10 15 20 25 30 40 50 60 70
Age in Years
Figure 6
Age-standardized defect size, age, and cauise of death in. 115 patients. A horizontal line is drawn
at 11 mm. to emiphasize that pulmonary hypertension due to ventricullar septal defect ha.s
never been reported to occur or develop in small defects.

are among the more common conditions that the belief that smaller defects,* i.e., groups
can be confused with defect. But few, if any, 1 and 2, by themselves, rarely if ever lead to
of the murmurs and signs of these diseases death as a result of changes in the pulmo-
are inclined to disappear with age and we are
1L
lett witn_ *L14
1 A -A
ot ')Z4A 1loud
-L 1 1_1L
murmurs, nalt wutn
__ -D1L
' The size is usually reported in linear dimensions,
thri'lls, that have disappeared in 24 years. not To age-standardize size it has been as-
as area.
sumed that any defect enlarges wvith age in propor-
Relk ition of Necropsy Material to the Natural tion to the linear enlargement of the heart. The author
Histtory of Ventricular Septal Defect adapted data representing the normal linear growth of
T he experience described above confirms the heart (pulmonary valve circumference) from two
sources 15 to make figure 7. Figure 7 is a plot of
the normal linear growth of the pulmonary valve cir-
cumference expressed as per cent of adult size versus
Qs )
90s age. To age-standardize a 5-mm. defect in a 2J2-year-
,)
old patient from figure 7, one finds that the 21,2-year
80 / point on the abscissa intercepts the curve at the ordi-
a3) c )
70
nate value of 50 per cent. The 5-mm. defect assumes
-U an age-standardized value of 10 mm.
> F:
60 / Others have followed the lead of Selzer10 and re-
a
lated size of the defect to that of the aorta. This has
c:
O merit, but aorta size is commonly omitted from autop-
O'_
:3
40 _/ sy literature concerning this lesion. Recent re-
ports 17. 52 (liscuss defect size in relation to body
30 _ 1 2 surface area. These data not only are difficult to cor-
15 20
Age in Years relate with older literature but also suffer from the
Figure 7 disadvantage that
defect size may be exaggerated in

Figure patients with growth retardation. Age-standardiza-

Putlynonary artery circumference (as per cent of adult tion of size, imperfect as it is, has the advantage of
size) versus age. These data are taken from two being applicable to all defects in which age and
.sour( es. rlimensions of the defect hav e been recorded.
Circulatioi Volurme XXIX, Jzne 1 964
 VENTRICULAR SEPTAL DEFECT
nary vasculature. If this is correct, then ne-
cropsy studies should show that only large de-
fects are found in patients dying with this
complication. Figure 6 is an adapted compila-
tion relating the age-standardized size of de-
fect to age and cause of death from 115 cases
taken in toto from several larger series in the
literature,9-13 in which data for individual le-
sions have been reported. To this compilation
have been added 16 cases where patency per-
sisted in an unselected series of 25 necropsied
patients over the age of 20 years who had
abnormal ventricular septa (discussed below)
and 22 further cases of all ages from the files
of 22,703 necropsies of Cleveland (Ohio)
Metropolitan General Hospital.* Figure 6 has
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
four symbols indicating the cause of death in
the individual case or the absence of informa-
tion: (1) deaths related or probably related
to pulmonary hypertension and congestive
heart failure due to left-to-right shunt (includ-
ing deaths following cardiac surgery when sur-
gery was indicated by pulmonary hypertensive
disease, and also including cases with minor
degrees of aortic insufficiency); (2) deaths
unrelated or doubtfully related to the causes
in 1, above (including cases of severe aortic
insufficiency); (3) cause of death not re-
corded; (4) cause of death, bacterial endo-
carditis. Difficulties are acknowledged in this
type of compilation. The size of a given defect
as reported in the literature is hazardous to
rely upon because of variations in shape, dif-
ferences in measurement technics, and status
of fixation. Digital preference for defect sizes
of 5? 10, 15, or 20 mm. is also evident. Finally,
the relation of complications of defect to death
in infants and young children is often difficult
to establish, and is therefore less significant
than are data from older age groups. Allowing
for these inaccuracies, there is unmistakable
evidence that smaller ventricular septal de-
fects never develop pulmonary hypertensive
complications.
*
There is a total of 31 patients with ventricular
septal defect in this necropsy series: four are included
in the series over 20 years of age; five cases, all
under 6 months of age, in which the size of the
defect was not recorded, were excluded.
Circulation, Volume XXIX, June 1964
929
It can be seen (fig. 6) in this selection of
115 necropsies that there is not one defect with
an age-standardized size less than 11 mm.
where the cause of death was related or prob-
ably related to hemodynamic complica-
tions. These data are in complete accord with
the follow-up studies of 8 to 12 years and 24
years, described above, and with the several
serial catheterization studies,16-9 including the
author's, which show minimal or no hemo-
dynainic changes in this group of patients
catheterized twice at intervals up to 12 years.
Additional data were acquired to prove this
point. The necropsy records of four general
hospitals representing 67,000 necropsies were
searched for cases of simple defect in patients
dying past the age of 20 and the size of each
defect was correlated with the cause of death.
Twenty-five cases were found (tables 12 and
13). The age of 20 was selected, since the data
concerning patients in groups 1 and 2 mo-
bilized thus far left no doubt that these per-
sons survive beyond 20 years (barring bac-
terial endocarditis).
Only five of the 25 cases were complicated
by pulmonary vascular disease, and the
smallest defect in this group measured 18 mm.
Four died of bacterial endocarditis and 16
of causes unrelated to their defects. Twelve
of the 25 patients showed evidence of com-
plete or partial spontaneous closure of their
defects (see below). Among eight patients
with patent defects who died from causes
other than pulmonary vascular complications,
the defect size varied from 3 to 12 mm. This is
conclusive evidence that progressive pulmo-
nary vascular disease for the practical pur-
poses never complicates the smaller defects.
Role of Spontaneous Closure in the Natural
History of Ventricular Septal Defect
(Tables 12-14)
For a number of years reports of sponta-
neous closure have been postulated or demon-
strated at necropsy.7 20-30 Rare necropsy re-
ports have usually described closure as being
caused by binding of the septal leaflet of the
tricuspid valve over the defect20 23 and re-
cently this has been described 24 in a case sub-
mitted for open-heart surgery following bac-
 930 BLOOMFIELD)
_0 z

;0 X
z
.Cw w
e

rv
J.
0 ._,
t
C)E) ow

S
C)' ~r _s

'' Sl ~-4 r
li
,t
0 _
01.
v

C
* ._
C) r. ,.
4-1
C) .
J1- t.

011 siv
4-- , _
X
0a t
X

01
rs .
E
o)
-2 -,'
_
.t , 1., Ct
0, _4
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

4--
C) ',i ;1 W

9
O U
U

C)
'-4

01-

r,1

4- C)

C)
it.

---C)
Ko. _ _
XC)
-X
C C
OOCO
in
c0 0C4
0
5

c,,- C0O0O
2, r 40 4C0 CO C0 C

- 1C . 1= L-i t Cs1 'r, CD1 C 00 t,>l 't Cico 'Iti't

c4 4-
C)01 ~~~~~~~~ccm
te O t CO t
It 41
t I_ C)

~ . i

~~~ ~ ~ ~ ~ >

-4 +

H- ~ ~ z1---~ z `

10
01 LC rl, I C^ Cn Z X
Cct 1l .1 oL
0,. --

01 01

0n1 O
C 5 '-t-- Int0 1n X
L-Co
C:1 t lt. t t r
-X

n c4
-Q
01 0>1 1. C0
1 t

1 c t l r,' 1,. n Z- I- 3Q
X
tn_ 4- L- 0C0 C-- - - ' -

Z.
;; Z;. ;; 1?-7
Z- z7 Z

C62 id2t!ion, VolZume.1 XXJX, jun(-c 1964

 VENTRICULAR SEPTAL DEFECT 931
terial endocarditis. Older literature has several
a3
a) a)
a
examples of this type of pathology, usually
4 c5
"a described as congenital aneurysm of the ven-
o C)
tricular septum.31-34 Plugging of the defect by
a)
v(IC
a4*8
C)a

.
a
vegetations of bacterial endocarditis 10 has
.a H o, c) $ a)
been observed. One report 30 describes the fus-
"5 , _,
X
C)= ,
c") C
,

ing of margins of a small defect in the

CS CS
>*E, ;,
c, muscular septum. Reports have also appeared
, .t 0 recently concerning the reduction or disap-
a) *_ pearance of left-to-right shunt or murmur in
younger children on the basis of clinical fol-
low-up or repeated cardiac catheterization
studies.22' 25-27 Hoffman et al.35 reported find-
-a)X
-
a
ings during open-heart surgery that describe
a) -CS W
CS
a) a defect squeezed closed during systole by hy-
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

c;t

Ct Table 13
z) a) Summary of Table 12. Isolated Ventricular Septal
St t Defect in Persons over 20 Years of Age Recorded
C) CSa~ at Necropsy
a
Per cent
a of all
No. cases
C)
C) CSO
C) Total cases 25 100
CS & W

Ct
tW
Cases with pulmonary vascular com-
CS
a " plications leading directly or indirect-
ly to death (minimum size of defect
at
CS
a
C
with pulmonary vascular hyperten-
o sion, 18 mm.) 5 20
h-a) Total cases showing a form of spon-
aH a
CS -
taneous closure, complete or partial 12* 48
C) _
- a Closure by medial leaf tricuspid
-
a
a $ )V^ a) valve, complete 7 28
Closure by medial leaf tricuspid
valve, incomplete 1 4
Fibrous patches or pouches 2 8
a) a)
Fistula-like tracts 3 12
Cases with interference of L - R
aD shunt due to aneurysm of sinus of
Valsalva 2 8
CS '- Cases of patent defect dying of causes
. *_I other than pulmonary vascular hyper-
tension or bacterial endocarditis (size
*a) range of defects: 3 to 10 mm.) 7 28
Cases dying with bacterial endocar-
a)
o >
a)
ditis presentt 6 24
a) Total deaths unrelated to defect 16+ 64
*
Case N-18, table 8, showed two independent
Cv aD
forms of spontaneous closure.
t In four of these cases death was considered due to
bacterial endocarditis; in two, the endocarditis was
believed to be an incidental finding.
|'- . '+ _ Includes two cases in which aortic insufficiency
due to prolapsed aortic valve cusp played a major
role in patient's death.
Circulation, Volume XX1X, June 1964
 9:32
pertrophied m-nuscle ar-otunid it. Animal studies
have also demonstrated spontaneous closture'
wlhethier niatuirally occurrinig or surgicallyCir-e
ated. Using dogs, Griffini anid Essex 'Ii showed
spointaneous healing within 2 mionths of surgi-
cally created defects up to 10 mnm. in dianmeter.
Others have shown'- that svynlthetic fiher plugs
with holes 3 inm. ini diameter placed in arti-
ficial defects created in clogs will occluide in
4 to 41 days.
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
92BLOOMNFIEII)
In a rev iewx of 67,000 necropsies at fouir
gen1er al lhospitals, the auithiorI fouinld 18 cases
wvi ere spontaneouis closi ire, coin-plete or in-
comiiplete, was possible or prolba1)le.2S Twelve
of these (among adults ovetr age 20) are pre-
senited in table 12. The remaining six pltus
three adc(litionail cases referred to the atlholr
are presented in table 14. Fou-r anatomic varin
eties of sponitaneouts closuire xvere, observed
aimon.?g tl-hese 21 eases: 1. Eleveni eases sl)owed

Figure 8
Occlusion of 17 x 10 m1i0. dlrfect by1 .septal leaflet of tricus.i,pid valve. A, left. Right cutticle.
B, right. Left ventricle (case N-2i5, table 12).

Figutre 9
lIucoruplete occlusion of 16-mmio. dfc/ct by septal leoifict of tricuspidi Al., left. Riglht vul7-
tricle. A black probe muay lbe seen entuergiug. fromn beneatlh te septal leaflet. Adjaceut to tIe
probe is an areal of curdocardial thlickeuirug. B, right. Left u utricle. (Case N-12, table 12.)
Cu uilton.l/ Vol/utne X'XIX, June 1964
 VENTRICULAR SEPTAL DEFECT 933

z 4).i
o o

~~~~~~ct
Cu

c)

a
.54 - C)
.=> : ;: 0 C)

_
C)
C) oCo)C
C CI o o
r.' ~ ~ ~ ~ ~
C)~~~~~P
c)

pp

c)

Z ZZZ Z C)
c 4 C'
.5i ct cc,- > z E-
C)C)4-

a -> cj-C v) +)
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

C)P

_ > CL) , o_ o

O
ct
C)

Ac

~ ~ ~ ~
~~t
C
QC)
I._

IC)
-C)

>
.O .^
CO Co CNI ~
X

_
uz .e
c
Co C0)](.

m
,E
Cirultin,Voum XXX Jn16
.C) H

O C 0 .*C O
CO
CC) 0
CC)
I!!r
COD
0
0) <C

00 in 0 C
^ ):
C]1 C>]

1Cn 0 C] C)~ 0
1I C]1 X--
C)C
V .

C)4

C-

cl

00 co
C]1 C]> Co Co Co

z Cz z z z z
 (-i34 B91i (0)0MFIELI)
Negro inan (ease N-31, table 14) xlxo dlied in
1958 follohi)w g surgical closture of a large
ostitino priimiutim, atrial s eptal defect comiipli-
cated l)v enidocardial fibrosis. In the ime-
l)ranolis septurm was the 2- by 4-mii. irregular-
ly (limpled sear shlowin in figure 14. The (lefect
was niot probe patent.
The demiloni.strationi of these aniomi-ialies at
postmorteni exaiiuniatioi does niot pro)ve tlexy
were ever patent (luiring life. Cases N-26 and
N-27 (table 14) are forms of ".spontaneous
clostur(e,' bl oth octicrrinlg inl the first year of
life. It may l)e argue(l that maly of the septal
antomalies of tables 12 anid 14 were present
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

at birtli, but in one case tlis alimIost certainly

was nlot so.
Figure 10
Left atsptricnlar ape.ct of case N-30, ttible 14. A 20
bij 25-nu2n7l. tr.anI,slaicetat fibwsous patch can be .scn i n
the ventricilar septun. (Photograph cou-rlt(esyl of Dr.
lleg(inltl hudson.)

l)inding of the septal leaflet of the tricuspid

valve. The dlefect was comipletely closed in
niniie cases (fig. 8) anid prole patent in two
cases (fig. 9). 2. Seven cases showed fibrouis
patchels of the muiscuilar venitricular septiuim
(figs. 10 to 13). 3. One case had an uiinuisual
dimpling of the memrbranouis septum siuggest-
inig postnatal closuire (fig. 14). 4. Three cases
liad residuial fistula-like tracts in the v entricu-
lar septuim surrounded by reactive tissu-e (fig.
15). These are forms that have been observed
oinly at inecropsy. Wlhat couild be a com-mon
form of postnatal sponitaneous closuire, i.e.,
closure of the membranotus septurm, woul.d be
dlifficuilt to trace at aultopsy. If one thliniks in
termns of a 1- to 9-mm. hlman meimbranous
(lefect closing spontaneously, it is difficult to
iimagi-ne how this wouild be observed at ne-
cropsy decades later. A fibrouis patchl within a
fibrous meml)ral-e woutld lhai.dly be recognized, Fi-ure 1 1
pa,irticuilar-ly sinec pathologists have inot been Left o ent title of c1 .3.3-year-old wtonic; it it/iadci/nc;-
alert to sulch a possibility. Figuire 14 is an im;e;;tet lhist/ory of muriri;;g nintupiur /)betitvet ape-,f's
(,xampl(. Thilis case was uncovered du rini1g a 26 cautd 33. 1 /1c I/olc in tihe snjperiol pontion of t1/i
musetiutiar septum has bleen coinplpt Ic/pely o/c/t/. (Case
casual examiniatioin of congeniital defects in. a N-33, table 14.) (Photograph cou itesy of Dr. Johnu
muiseuim collection. This was a 15-year-old Cria.)
Cairala/ion, Vulmine XXIX, Jimne 1964
 \VENTRILCULAR SE1PTAt.L DEFECT 9'35

C(asw IIe poirt.* A :3:3-xvea,r-old motbier oftxo

(ease N $33, table 14) died of staipbl-ve)occa]i
pnieimoniai and sel)sis at ter.m pregnaneyw She
had. the phy sieal. liiidIi igs Ut a mild (gr.oop 1-2)
lesioni unitil age 26, whien the. mutrmur and signs...............
disappeared. She was \\elliiuntil hier fatal illniess,..
wheni shec was adm-itted to the Boston Lyinig-in
Hospital tor her third cesareani sec.tion after- ani
uniieventfu-l pregnancy. Uponi entry the patienit had.....................
a iow-grade fever and signis of an u-pper respira
torv infectinni. Anemia (hernatocrit Ieve]., 28)
was observed for, the fir-st timie. Shie died 6 clays
later. At inecropsv the hiear.t weighed 400 Gm. anid
was diffuisely enilar-ged. The left venitricle was 2.0
em. thick; the right, 0.6 cm. Sectionis of the hear.t
(fig. .13) showed. a genieralized hyvpertr-ophv~of
mu-scle. Thetie xvas a "closed scarred foramen-like
depressioni 1 cm. inl diameter .. in the uipper
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

.......

I am indebted to Dr. A. XV. Tuicker of Boston,

Massachuiisetts, for permission to report this case; to
1Dr. Joh-n NI. Craig, Paithologist-in Ch"lief, Bostoni [y-
ing in Hospital, for the pathologic find(-ings, photo-
graph ( fig. 1.1) anid sections across this clefect ( figs. '
12 and 1:3); and to Dr. Cuit tis Prouit, Dedhiam, Massa-...
chuisetts, wxho very kindcly described the early hiistory
of this 1)atient. Figure 13
High-4j owcer i)i-thotinicrogriaph1 of vetrnicidatr p~.1tal
ninsete shiowEing mnlocardial hqpllertrophqit. (Phiotog,raphl
collrte,st/, of Dr.. jolhn Craig.)
third of the initersvenitricuilar septu-m" (figs. 11.
and. 12). This patien-t is the on-ly case of niecropsv-
~~~~~~~prov ed sponitaneouis closure in this ser.ies, or any
........... (~~ther kniowni to the authtlor, xxith an adequiate hiis-
...tory of a mu-rmuir th-at disappeared. Closiit e
~~~~~~must have occurr-ed betxxeen the txxenity-sixthi and
~~~~~thirty third year. Th.e pathology is xvery similai.
...to the case report bx' Roberts et al.,9 i xxhc

A teondcasv o poabe spotne ous

.losurein.anadul
....folloxv.up.study. oidccured
............ in the 194i 96

... ... etreiommesndation -ihis certificae readn,"er

...enlarged....one. and..one. half..inches;...bruits...heard.
...

......ovr.all aras..Markd.dyspneaon.exerton.
.able.to do.duty ..For.the.next 20.years.he.took
Neve
Fiue1 ..ligt.jos.bu.coninue.tocompain.f.brath
eses exa efre oteNtoa er
Low-po..r...
ter.. photo...ro..ra... ..r.oss rtcf. t offgr.Hsia.vihtee.opans.n13.a g 6
....I..i..(Phtogap con..
s..i...of.D....i.......... .o..C a ).The.clinical findings at.this.time.were."Heart.not

Circsulaton. Volume XXIX. June I96*

 936 BLOONMFIELD
ia1erged. Louid pirecordial sy stolic m1ntmiir.
_tlli-int -I ill aii pansx stolic nlllllimri. at the apex xx ti
*-otetl. An eletroeartl)gan is sllolxxii i figue
gi
16 left. Oi floiosco examillatiot thelleeart oas
ormal it aize, Ntit theire wxas p()ii mtncnt o
th.e puilmonaiy ar~terv. A tentatixve diagnosis of
ventricilar septal defect or piilnuixiltv steinosis
aZs m1ade. Thei patient wxas. stat tedl oni digitalis
for- the first time, l)ut xxas thien lost to follow uip.
He has remained xell for- the past 2:3 vealrs. Dtir-
toig thi8s periodl, because of the heart mur.muri.,hle
xxas txxvice rejected as a life inisuran-ice iisk. When
traced in 1.960, at age 69, hte piesieted himii 4sell
as feeling "better than e IHe still complainedl
of mild exet tional dy spnea. He had never' taken1
digitalis except for the sh
Q..E -~~~~n197Th)001pessuiie xx-as 150, 70, puilse 72.
The patient xxas a slighit buit normal appearing
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

Icmlt mcian ofl 69. Thete xxas prominence of his left

chest in the r-egion. of the puilmonatrx artetry. The
aex hea.t xxais nior-mal. Thii. erexccitto abnoorma!
cariacliie pulsations o thirills. Oni autscxiltationi
(plhonocardiogzram, fig. xx17). followxinig a1 noimal
fist heart sonnd there as a loudc ejectoiot click
that diminished slightly x t insplirationi. The
..~ ~~ ~ ~ ~ ~~~~~ite ....l
secoti liiisi
st.i5Ointcl -o-t
in eillg
the ettlitarea
puilmionaiv lesmsf
xxaslilpet-
trL
sistentlv split anti xariedl nor1mallv xxith respira-
tion. A st)ft aitrial souintl xx as atidlible. There xx as
no c itlence t)f the precortdial. sxystolic murimuti
th.at hiad beeni tdocuimentedi 23 aind 44 years bwfore.
FigureFigure
14 ~~~~~~~The,
electrocartdiogram (fig. 16, i ighit) was un11
14

Lft trent 'iuar Isp)tX of dinipled, nOPtt(tire

~~~~chianged
from 1937. Ch-est x-rax (kintily ititer-pret-
eti ]) Dr. Peter. Kerlexv) shoxxeti a 1argc' puilmo-
ollar septa I tlefct'. (Cose
.31, taIl(' 14.)nary arterx anti a small aorta (fig. 18). Dr.,
Keriev, xxho xxas not iiufoi-med of the clinical
history pr-ior- to hiis ieading, felt thazt the most like-

.Figure.1
Incomlete ccluson ofueuitWolispa eetscnayt olonto cs.A et
Redutvcntoh B n ht 1 cft cuiti tub (Case N-20, table 12.)..........

(Cl}#i a aton. tunice XXVLX, June 1964

 1v
or
the

u:qnWigetV4
VENTRICULAR SEPTAL DEFECT

(liagniosis
puilmonar a:terr
l)ecause
asat suirgically closed veintr-ictlular
aitriall septal defect. Idiopathic dilatation of
of the large
possible to prov e thalt thlis case
xx as
size of
con sidered less
1)0th the left a-nldl
right ptilmonaryi artery branches. Wlile it is in-s
w ever
likelx

as a
eiitricuhnl septal dlefect, it is
recoicile this cotur-se
tiooi. Tlhe disappearance
uni iqItue helest
(Id fisieat on
x-i av
nith

speak
of

exe-!rtioii malv be
it stalted durinig XWorld
aln1d has h)eeni uniclhanigedl for the past
jiist as

the imurmiiiui r
fo- themselxves.
nietur.astlenic,
\\War I militarv
[She electr-ocard(liogl am i's typical, and, since the
cha.inges aire

heart disease.
niot diue to digitalis, it stiggests
obscuire left xventrieular mvTopatlhv of

A case of incom-plete sp1ontaielouis closure

in an adtult is diseussed l)ecause it provides
evidence that inflammnation mnay iniitiate the
closuire proc ess.
benigni
varioty (see belox ) Tli-ere is nio clin-ical basis to
thinlk that the abnormalities are (Iide to iselhemic
difficult to
othler kiiooxva coindli-
and

in
seivice
44 years.

a
937

the
The
that

an
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

T:.
mffli.VV6w Ca.se) Re,port. Thlils 61-yveari-old1 mnani (case N-20,
table 12, fig. 15) died slhiortly after his first ad-
mission] to University Hospitals, C(leXvelantid, ini
1942, xith the cliniica,l dlia,gniosis of lhvper-tenisive
Figure 16
lhear-t disease, conigestive heart failure, aniid uiremia.
At necr-opsv the unuisual defect and vegetative
Electroccdcliogrcams cit ciges 46 (left) and 69 ( eentie enidoc,ai-ditis of the rlighit xventriele (fig. IWA)
antid right) of ci man. with1 clinical history sting xxere fotunid. The defect xxvas descrihed ---a round as

clostsize of vcenotricular septal clefecct ai

,sionitcinccsns dlefect, 3 mm. in diameter, surri otundcled firm
some tinme during this periiocl. g rev tissue in a shallow- fuiniel weichprotrudes

PA ic

MA
XINS P.

Figure 17
Phonocardiogram of .samce patient as~in figure 16 cit age 69. A prominent eic'ction click is re-
corded. PA, puhnoncry circa; LSI, loiter Lift .sterncial edge: MA, miitr.al area; C, ejection click
occoirrinig 0.05 sec. caftcr thce first sonncil.
( rid atiou, Vnlmu,,e XkMI 5. Jimnic 196-s
 9,38 9:35
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
Figure 18
Chest x-/ail of sanwc patitnt (as in fgu16 (it. age (5
as a pouting mnass leilgthe veiitijela,tii sep-
tal dlefect In the righit xeit-entriclzn aspect. lRadlat-
inlg aice hard masses of calcfium to the sinuils
of Valsalva, aortic xvaixve, imighit xventiricu.lr xv all
and l)ehid the auitemi-oi, leaf of the timicuispicl
Vaixve o)1 sec.tioni theme wxas massivx me-
I)laLcWment of mfislell Iy co~llageni and coni
nective tissuie xxithi chr1onie inflammation.-- The
pzatienit hanl dlen-ied knioxxlleclge of miruir-ii-s in his,
earilv life, huit hiis fatal illniess r-epr.esen.tedl his Li.ist
conIta,.ct xxiha do(ctoi This may ie(~pme.sent a Conl.
genital defect in the piO(Cess of closure stimnlated
In, ani iniflammaitorxv reslpose.
Tihere is nio doubt that sponitaneou~-)s closure.
occulrs. Tihe poinits at issuec are ( 1) f-low often
does this happen? (2) To whiat ty pe of (le~feect?
and ( 3 ) At xx'hat age (ioes the ex ent commoiinoly
take place?~TImec evidence fromn chiildren)'s hios-
pitalS 22' 2'' suggests that closure oeccurs in
perhlaps as miany as 25 per cenit of dIefects ini
the. first fcxx years of life. Evanis et al.22 ob-_
ser-ved possihie closuire ini 37 of 120 (31 per-
cenlt) patienits xvithi a par-ticuilar type of ds
appearinig mutrmuir that enide,d hefore the--
seconid sounid. This type, of muixrmuitr ~md dis-
appearfing dlefect was fouind in a gr-oup of 5-4
of 2177 (25 per cent) snibjects wi-th x'entricular
septal defect. Eighity per cenit closed hy age 3.
Nadas ct al.'2" observed hiemodyviana ic proof of
sponitaneouis closuire in fouir cases, thiree of
whichl h-ad puilmiona.ry hyipertenisioni close to
system-ic lexvels. Agustsson et al.2' noted closuire
or ameliorati.on ini ei(lrbt
C) of 1-95 cathieterized
131 f~~~~~~~~~L0O\1F11ELD
cases (4 per cent). In hlalf' of these the def ects
causd crdioegax ,Pttliioia- plethora, and
left xventrictilar hyxpertrophyx In our 24-sear
follox Ii) stdx. ighlt ot 11 patients xxit
smlaller df ceets tin1der II1 years of age and si\x
of 10 pi~tienits 11 y ears annd older lost their
mnurninurs. Exvcn if xxve assumn-e that the diagniosis
xxas correct in onlyxlhalf the c.ases, this would
still iml-pix that as mianyv as seven of 21 cases
closed sponitaneon,)isix in the 24-vear per-iod.
In summiiiarxv, inirciiect cliniical cxvidence leadis to
the conclusion that perhaps 25 per cent or
more1 of small defects in children surxviving
infiancv xxviil lclose spon-taneion sixv, gi nthe op-
portuniity. Onec additional evenit that iioist be
considered is par-tial closure of a large diefect
(fig. 9 ), folloxx ccld by complete closuire of what
\v as thonighit to he a smnall defect.
Other evidlence hears iipon thiese coit-
clusions. First, in the 12 yer of exper-ience
of I)r. M ood and the National Ileart Hospital
xvithi cardiac catheteriv'ati on s ino catlheter-
prov ed case, hoxxever mild, hia5 been observed
to close. Second, anix tlsis of incidence _fignres
and antopsy dlata suggests that the incidence
of (lelects appearing at autopsy is close to
xxhtmight he expected. This statement is
based ipmoni ( 1) the incidence of congeniital
heart dlisease andl x entricular septal diefect inl
patients surxvixviiig inifaincy ( i.e., ini unselected
ptopullations of school chiildren'-''') ( 2) the
assuimptionl that mi-ost mlild defects survive. he
xond 1,5 or 20 xears; and (3) the. analysis of
incidence at autopsx in patients oxver 15 or 20
x cirMS of age comipared xxith xx hat miighit he
expected froin statemenii,its I and 2, above.
The mi-ost reliable figures place the incidence
of conglenital hecart di-sease ili patients survix-
ing infancy betxxecu 10 and.e 20 per. 10,000
peirsulis. \It\lac.~lioui et al. in Liglganid p)latc
the incidence of congenital heart disease at
age 10 as 10 per 10,000. Glardiner anid Keith, 41
in Canaii~da, calcuilate the incidlence in schlool
children at 21 per 10,000. Rliauh" in the
Uniited States noted an incidence in schiool
children of 15 pcir 10,000; anid Bobinson cet a L:I."
nlotedl 19 per 10,000. Otheri axvailahle figuires
geuleradlv confirm- these aesults. It seems rca-
sonable to conclude that the inicinlen.ce of con-
al- 17.o'!iiimc \XI\ Junca 90
 VENTRICULAR SEPTAL DEFECT
genital heart disease in children of school age
is close to 16 per 10,000. Ventricular septal de-
fect made up 38 per cent,38 34 per cent,39 36
per cent,40 and 36 per cent41 of cases in these
studies, an average of 36 per cent. The ex-
pected incidence of simple defect in an un-
selected series of autopsies of persons of school
age and beyond would therefore be 5.8 per
10,000. This figure is probably high, for Dr.
Wood has pointed out that the diagnosis of
simple defect in school children is frequently
confused with pulmonary and aortic stenosis,
mitral insufficiency, and the acyanotic form of
Fallot's tetralogy. If one assumes a diagnostic
accuracy of 50 per cent at the time these sur-
veys were done, the expected necropsy inci-
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
dence is modified to lie between 2.9 and 5.8
cases per 10,000. This is approximately the in-
cidence actually found.
Selzer,10 in reviewing the literature, found
30 cases over age 15 among 45,894 presumably
unselected autopsies,* a rate of 6.5 per 10,000.
Among the 67,000 autopsies reviewed by the
author, 17 examples of patency occurred in
adults over age 20, a rate of 2.5 per 10,000.
Fontana and Edwards42 found but a single
case of ventricular septal defect among 22,361
postmortem examinations. Kaplan et al.,43 re-
porting a smaller series, found four cases in
6,983 consecutive adult autopsies. Therefore,
a total of 52 cases has been found in persons
over 15 or 20 years of age among 142,000 con-
secutive autopsies, an incidence of 3.7 per
10,000.
This does not complete the reasoning, how-
ever, for it is tempered by the experience of
our 24-year follow-up study together with cur-
rent clinical experience. Only two of the 48
cases (4.2 per cent) found in reviewing the
National Heart Hospital 1934-1939 records
could have been in groups 3, 4, and 5. We
know from our present series, when diagnostic
technics have improved considerably, that
35.6 per cent of all cases (table 1) fall into
groups 3 4, or 5. Presumably those with pul-
monary hypertension were often confused
* Among his own personally observed series Sel-
zer 10 found only three adults among 7,243 autopsies,
a rate of about 4 per 10,000.
Cifculation, Volume XXIX, June 1964
939
with rlheumnatic hleart disease, a far more
prevalent diagnosis at that time. Since the
figures on incidence in school children were
gathered largely before adequate diagnostic
technics were available, we must assume that
the calculated incidence of ventricular septal
defects (2.9 to 5.8 per 10,000) represents
mainly the smaller defects with more classical
physical findings. If we correct the school-chil-
dren incidence figure of 2.9 to 5.8 per 10,000
for the few larger defects that might have
been included (4.2 per cent), the incidence
of small ones (groups 1 and 2) in school
children and therefore at autopsy in adults
should lie between 2.8 and 5.6 cases per
10,000.
Selzer's incidence of small defects (less than
10 mm.) in patients over 15 years was 20
per 45,894. In the author's series, 10 cases of
patent small defects were found in 67,000
autopsies.* The observed incidence of small
lesions found in adults at autopsy for these
two series totals 30 per 113,000 or 2.7 per
10,000.
In conclusion, if one accepts the diagnosis
as correct in the school children studied, then
necropsy incidence implies that spontaneous
closure occurs in 50 per cent of small defects.
If Dr. Wood's estimate that the diagnosis of
defect in school children is in error in half
the cases, then spontaneous closure would ap-
pear to be an insignificant factor in the natural
history of small defect. In the author's necrop-
sy series (table 12), eight of 25 cases in per-
sons over 20 showed evidence of complete
spontaneous closure (32 per cent). On the
basis of the clinical evidence cited above, and
a compromise value from autopsy evidence,
the author believes it reasonable to assume
that at least 25 per cent of small defects
among patients surviving infancy will close in
a lifetime.
If the 25-per cent figure is correct, the in-
cidence of spontaneous closure that should be
found at autopsy would equal one third the
observed incidence of patent, small defects, or
* No other large series of unselected autopsies in
which age and size of defect are recorded are known
to the author.
 940
0.9 cases per 10,000. Again from the 67,000
cases cited in table 8, the eight cases of spon-
taneous closure cited in these adults give an
incidence of 1.2 cases per 10,000. The esti-
mate of a spontaneous closure rate of at least
25 per cent not only appears valid but also
suggests that most cases closing spontaneously
leave traces that competent pathologists will
observe if they look for them.
Finally, attention is called to the obstruc-
tion of ventricular septal defect by a potential-
ly dangerous aneurysm of the sinus of Val-
salva (cases N-11 and N-15, table 12). This
complication should be considered in any case
of "spontaneous closure,"' complete or incom-
plete. Case SC-2, table 8 (case N-li, table
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
12) is an example in point. This 41-year-old
woman had cardiac catheterization shortly
before her death. The data showed normal
pressures in the right heart and a very small
left-to-right shunt. At autopsy her 12-mm. de-
fect was partially obstructed by an aneurysm
of the sinus of Valsalva. Death was due to
rupture of the aneurysm from bacterial endo-
carditis.
Lack of Relation of Idiopathic Dilatation
of the Main Pulmonary Artery to Spontaneous
Remission of Left-to-Right Shunt
In view of the findings described above, it
seemed possible that spontaneous closure may
account for some cases of idiopathic dilatation
Table 15
Clinical Findings in 36 Cases of Idiopathic Dilatation of Pulmonary Artery and Its
Lack of Significant Relationship to Possible Spontaneous Closure of Ventricular Septal
Defect
Total cases
Age
0-15
16-30
>30
Male
Female
Positive history of disappearance of heart murmur
History possibly consistent with previous heart disease
Pulmonary ejection murmur
Pulmonary ejection click
* This series is highly selective for males of military age because of referral due to abnormal
chest x-rays at time of entry to military service.
t See text for discussion.
BLOOMFIELD
of the pulmonary artery.44 Thirty-six patients
with this diagnosis were listed in the files of
the National Heart Hospital. Approximately
75 per cent were in young men referred to the
hospital because of an abnormal chest x-ray
found in routine screening for military service.
Diagnosis was made on the basis of an en-
larged pulmonary artery segment by x-ray
without other evidence of heart or pulmonary
disease. Electrocardiograms were normal. A
soft pulmonary ejection murmur was heard in
42 per cent of cases and a pulmonary ejection
click in 39 per cent. The second sound was
physiologically split, and the pulmonary com-
ponent was not accentuated.
In this retrospective study each case record
was searched for evidence of prior heart dis-
ease, disappearing murmurs, or changes in
symptomatology. Very little was found to sug-
gest that idiopathic dilatation of the pulmo-
nary artery was related to spontaneous closure
of ventricular septal defect. These results are
summarized in table 15. Only two of 36 pa-
tients gave a history even modestly consistent
with the disappearance of a heart murmur.
One was an 18-year-old boy who gave a his-
tory of "rheumatic fever" at 9 months of age
and again at age 14 when he had "growing
pains." He was unable to state whether any
murmurs were heard by his physician at these
times. When seen at the hospital he com-
No.
36 100
3 8
27* 75
(3 17
26* 72
10 28
0 0
2t 6
15 42
14 39
(irculation. Volume XXIX, Juine 1964
 VENTRICULAR SEPTAL DEFECT 941

Table 16
Incidence of Bacterial Endocarditis in Ventricular Septal Defect
1948-1960 1934-1960
1948-1960 25-patient 48-patient
NHH patients study study
Total cases ventricular septal defect 170* 21t 8t
Total cases bacterial endbear-
ditis 6 1 0

Total deaths due to bacterial

endocarditis 0 1 it
Total patient-years in follow-up
Minimum 222 1711T
Maximum 360** 259tt
Cases of bacterial endocarditis
occurring during follow-up 1 1
Deaths due to bacterial endocar- 1
ditis during follow-up 0 i+t
Risk of bacterial endocarditis;
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

cases per 70 patient-years

Minimum risk 0.20** 0.27tt
Maximum risk 0.32 0.41
Mean risk 0.26 0.34
* Excludes National Heart Hospital patients in 8- to 12-year follow-up study.
t Includes only patients located in 1960 with ventricular septal defect (alive and dead).
+ Occurred in pre-antibiotic era. Patient had bacterial en(locarditis when first seen in 1937.
Based on 21 patients from time of catheter to time of follow-up in 1960 or death.
T Based only on seven patients located in 1960 who still had defect. Excludes patient de-
scribed in note t since bacterial endocarditis was present when first seen.
** Based on 25 patients selected for follow-up; from first visit to National Heart Hospital in
1960.
tt Based on potential of 23 patients, located, assuming 50 per cent had defect; excludes
five patients with wrong diagnosis and patient in note t above.
++ Group-4 patient who died at age 8 of aortic insufficiency secondary to bacterial endocar-
ditis originating in the defect.
Ages at which endocarditis occurred in nine cases were 8, 19, 19, 28, 29, 30, 30, 33, and
53. Five cases were male, four female. Eight cases were in groups 1 and 2, and one in group 4.

plained of mild dyspnea on exertion. The only
abnormal finding in this patient was pulmo-
nary artery dilatation in the x-ray. The second
patient, a 35-year-old man, gave a more posi-
tive history of "a bad hearf" at age 15. He
was accepted for military service at age 21,
apparently free from heart disease at this time.
When he was seen at age 35, aside from pul-
monary artery dilatation a pulmonary ejection
click was heard.
Risk of Bacterial Endocarditis in Isolated
Ventricular Septal Defect
Bacterial endocarditis is a major risk con-
fronting patients with defects. Pathologic re-
ports leave no doubt of this. Gelfman and Le-
vine45 found bacterial endocarditis in 42 per
cent of 31 autopsied cases of defect over the
Circulation, Volume XXIX, June 1964
age of 2. Selzer 10 found it in 26 per cent over
the age of 15. The data in table 12 show it in
five of 17 patent defects over the age of 20 (28
per cent), excluding one case with infected
aneurysm of the sinus of Valsalva. One may
question how valid the autopsy figures are. Do
factors of selection exaggerate the risk? What
is the risk per patient-year, or better yet, what
is the risk per 70 patient-years, a normal life-
time?
The risk of bacterial endocarditis is best
determined by complete follow-up of a group
of patients for an arbitrary period. Risk is cal-
culated by dividing total patient-years by the
number of cases occurring during the period
of observation. The follow-up data from the
12- and 24-year studies have been analyzed
 942
in this fashion and are summrniiiarizedl in. table
16. Difficulties are immlediately aIpparent. The
incidence of lac terial endocarditis within
these controlled periods of clinical obserxvation
was very sinall. In the total National Heart
Hospital experience between 1948 and 1960,
and including the 48-patient study, onily nine
cases xvere recorded; five of these were pa-
tienits who, wlhen first seen, had bacterial endo-
cairditis onily by past hiistory. Aothler pre-
senited becau.se of symptoms fromn bacterial
enldocarditis (in 1937), and onily three cases
occurred, to ouir knowledge, after their first
visit, i.e., d(lring a periotd of follow-up.:t Dr.
Wood's personal experience of 638 patient-
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
942O131
X)OM FiELD
Cardiomyopathy of Ventricular Septal Defect
It is con-nmonly assuniel that patients withl
a small defect, inaladie Ie Roger, have a nor-
nmal lite expectancy, barring infective endo-
cardliti.s. For two reasons this is probably niot
truie. 1. As previously reported,', the author
ha.s collected nine eases of minior defects of
the ventricuilar septuim in tlhree living ancd si\
inecr-opsied 1)atienrts in whoim there was evi-
clenice of left veintrictilair mnyopathlic chanige. 2.
The patienit wxith a small defeet whvo suirvives
beyond the age of 50 uncommonly presenits in
coiigestiv e heart failure. This stiggests that
the trivial defect of youith begins to take its
toll- as the heart ages or in the presenlce of aiy

years of ventricular septal defect incluided not ui-nrelated conditioni that compromises cardiac
a single case of inewly developed bacterial ftunction, e.g., hypertension or coronary artery
endocarditis. disease. SeIzer 1 certaiinly had a presentiment
What remains are two cases of lacterial that this xvas so, for he commented oIn the fre-
endocarditis oceitrring duiring the controlled uti licyvf coungestive heart fatiluire in his series.
follow-uip periods, one eachl in the. two studies. He noted that six of 35 patients over 15 (17
The authlor makes nio claim to statistical p.er cent) died in congestive heart fatiluire pre-
validity, but simply observes firom table 16
that the mean risk per 70 patient-years is 30
per cenlt. This figuire is rather close to that of
the auitopsy sttudy material cited above. It
stuggests that the atutopsy material from un-
selected souirces is representative, anid not se-
lective for bacterial endocarditis. Furthermore,
eight of the nime eases in table 16 occurred
in groups 1 and 2, i.e., the grotups most likely
to survive to age 70.
All three cases of l)acterial encdocarditis that
occurred dturinig the follow-uip periods lhap-
pened within the antibiotic era. One of these
died, the grouip-4 patient. Since bacterial en-
docarditis in ventricuilar septal defect is now
frequently cured witlhotit sequielae, the mortal
risk is probably only onle fifth the infection
risk. The mortal risk for bacterial eudocarditis
in this era is then 6 per cent. Corrected for
spoontaneous closuire, the figuire for lifetine
mortal risk falls to soinething beloxv 5 per cent.

F[he important contribution of the Salxation Armv

workers to this stud(ly canr he appreciaitedxwhen it is
realized that two of tlhe three cases of bacterial ciido-
carditis in the 12- and 24-year stti(lies xx erc folnd
ollly becauise of their efforts.
Figure 19
Chlest x-ray of 52-y)eifarold mizan wcithl small defect a(1
early left teutricidar failurle. Similarity to patient of
figurei'c -8i8s apparent. (CascSC-3, table S.)
Circultu/on, V/Oline M/XXKX, fune 1964
 VENTRICULAR SEPTAL DEFECT 943

taken cardiac medication such as digitalis. Physi-

cal examination revealed a blood pressure of

I 110/60, a pulse of 72, and normal jugular venous

pulse. His chest was "pear" shaped with flaring of
the lower ribs and depressed diaphragm. The
cardiac impulse was normal. There was no thrill.
On auscultation, at the left sternal edge im-
IT
I mediately following a normally split first sound
there was a grade-II/IV decrescendo murmur end-
ing well before the second sound. The second
V4 sound in the pulmonary area was normally split
and not accentuated. An expiratory third sound
was heard at the apex after exercise. Cardiac
catheterization studies are shown in table 8 (case
SC-3). The findings were consistent with the
diagnosis of a small defect with slightly elevated
pulmonary vascular resistance (2.6 units). The

m elevated pulmonary vascular resistance was proba-

Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

bly a passive phenomenon, since pulmonary

wedge pressure was high normal at rest, rising
Figure 20 to abnormal levels during exercise, and cardiac
output failed to rise with exercise. Pulmonary
Electrocardiogram of same patient as in figure 19. vascular resistance also rose with exercise. These
Note similarity to electrocardiograms in figures 16, 21, findings are consistent with left ventricular myo-
and 22. pathic disease. The unusual shape of the chest,
the emphysematous changes, and the arterial de-
sumably unrelated to the shunt. Two had saturation, which felil with exercise, suggest an
hypertension. additional element of primary pulmonary disease.
Among the nine cases collected by the au-
thor, attention was drawn to the problem of
cardiomyopathy and survival by the second
case described under spontaneous closure
(above). The electrocardiograms (fig. 16) I V1
were abnormal in 1937 and in 1960. The intra-
ventricular conduction defect, ST-segment de-
pression and flat or biphasic T waves in II,
III, and V4-V6, are consistent with nonspecific
left ventricular disease. The changes are re-
II
markably stable, not due to digitalis, and not
at all associated with coronary artery disease
symptomatology. The patient denied previous
diphtlieria. At 69 he was well, presumably
with a closed defect. Shortly after this patient
was seen in 1960, a second case appeared at
VV5
the hospital, symptomatic, with a small defect,
and a chest x-ray (fig. 19) and electrocardi-
ogram (fig. 20) quite similar to the first. VR vT
Case Report. R. M., a 52-year-old man (case
SC-3, table 8), was seen for the first time in 1960 Figure 21
complaining of palpitations and mild dyspnea on
exertion for 18 months. He had had a known heart Electrocardiogram of 15-year-old boy with defect
murmur since his early teens but was never fol- shown in figure 10. An uncomplicated atrial septal
lowed regularly by a physician. He had never defect-secundum was also present.
Circulation, Volume XXIX, June 1964
 VB,3mnI!1mlLM
944 BLOOMFIELD

Case Report. K. F., a 15-year-old boy (case 3),j
was first hospitalized for "primary tuberculosis"
between ages 6 and 8. Review of x-rays from this
period suggests that the diffuse infiltrate thought
to be tuberculosis was actually pulmonary plethora
associated with a left-to-right shunt. At no time
during this hospitalization was a cardiac murmur
noted. When the patient was seen at St. George's
Hospital, London, in 1959, physical examination,
chest x-ray, and cardiac catheterization were en-
tirely consistent with the diagnosis of atrial septal
defect, ostium secundum. The electrocardiogram
(fig. 21), which has features similar to cases 1
and 2, was puzzling and suggested the diagnosis
of ostium primum atrial septal defect even though
the T-wave changes in the left precordial leads
were unusual for that condition in the absence of
mitral insufficiency. The patient was selected for
closure under hypothermia. At surgery a small,
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
Among living patients, there is one more
case to be described. A group-i woman of the
8- to 12-year follow-up series had normal
electrocardiograms at ages 27 and 33. At age
36, while the electrocardiogram was still with-
in normal limits, early ST and T-wave changes
in the left precordial leads appeared (fig. 22).
In summary, nine cases have been observed
in which there is an association of a cardio-
myopathy with defect, actual or probable,
open or closed. Two of these 10 are catheter-
proved cases from our current series and re-
I._

ordinary secundum defect was seen and closed.

The patient died of a hemorrhagic complication.
At necropsy (case N-30, table 14), in the mid-por- vF
tion of the ventricular septum there was a thin,
translucent, fibrous patch 20 by 25 mm. with a TIIJI II
dimpled center (fig. 10). It was not possible to
determine whether this patch represented a con-
genital defect in only the musele of the septum or
whether it represented the postnatal closure of a
patent defect. The central dimple suggests a con-
centric closure. The left ventricle was thickened
and hypertrophied.
The fourth case in this series was the 33-
year-old woman whose history is described
above as an example of spontaneous closure {fe1F '-11-1-0;
(figs. 11-13). Although she was pregnant and
|l]T0L t
anemic in her last trimester, the generalized
cardiomegaly (heart weight 400 Gm.) was
unexpected. Electrocardiogram was not taken.
From the necropsy data in tables 12 and 14,
four other cases are consistent with a defect
and left ventricular myopathy. Case N-13 was
a 45-year-old man who died of bronchopneu-
monia, in 1925, and had a small 3-mm. defect.
V44 S ;
His heart weighed 400 Gm., the right ven-
tricle was recorded as 2 mm. thick, the left .-- t
2ii
ventricle 12 mm. Cases N-24 and N-25 both
showed nonspecific left ventricular hyper- V5 1. 1 .. ..

trophy, although clinical history was available .-.t

in neither. Case N-31 (fig. 14) had left ven-
tricular hypertrophy and endocardial thick-
ening demonstrated at necropsy.
* I am grateful to Mr. C. E. Drew and Dr,
tham for permission to report this case.
Figure 22
Serial electrocardiograms of a group-i woman at ages
27 (left), 33 (center), and 36 (right). The last shows
early ST and T-wave changes in the left precordial
A. Lea- leads suggesting the onset of left ventric.ular cardio-
mnyopathy.
Circulation, Volume XXIX, June 1964
 VENTRICULAR SEPTAL DEFECT
present 10 per cent of persons over the age of
30. Neither electrocardiogram nor myopathic
changes were found in any of the other older
patients in the National Heart Hospital series.
No literature exists on this subject but the au-
thor noted that the second case described in a
recent report47 bears all the earmarks of the
first two myopathy cases described above, al-
though it was reported for a different reason.
Turning to necropsy material (table 12),
six of seven patients dying over the age of
50 had some degree of left ventricular hyper-
trophy. Only one of these had myocardial in-
faretion. Four of seven died of congestive
failure ascribed to hypertensive cardiovascular
disease, and a fifth was known to have hyper-
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
tension but died from pulmonary embolism
following an automobile accident.
There is no other evidence to support the
concept that systemic hypertension is more
common with ventricular septal defect. Fig-
ures from other series are not available. What
is likely, however, is that a defective ven-
tricular septum, patent or not, compromises
cardiac function when the heart is subjected to
a stress such as hypertension. Since four of the
seven cases over age 50 were examples of
spontaneous closure, it appears that the defec-
tive septum per se, not the small shunt, is re-
sponsible for the increased susceptibility to
cardiac failure. This is important, for it implies
that this long-term complication will not be
influenced by surgical closure of small defects.
It may be that many of the patients with
large defects now closed surgically will en-
counter this problem after living to ages pre-
viously unattainable.
One can only speculate upon the etiology of
this myopathy. Most likely, a defect of the
ventricular septum mechanically interferes
with systolic contraction, i.e., the ventricular
septum cannot make its normal contribution to
systolic ejection. The result is lateral wall
strain as indicated from the electrocardio-
grams in figures 18 and 20 to 22. This, too, has
suirgical implications, for it is entirely possible
that synthetic fabric patches now used to close
many defeets will stiffen with age, sensitizing
patients to early left ventricular failure.
Circulation, Volume XXIX, June 1964
945
Brock48 has used double-thickness pericardial
grafts to close larger defects; this approach
may prove to be a treatment of choice over
the years.
Origin of Increased Pulmonary Vascular
Resistance and Pulmonary Hypertension
in Ventricular Septal Defect
The origin of pulmonary hypertension and
elevated pulmonary vascular resistance is still
in debate.49 It cannot be explained on the basis
of any single factor. The change in physiology
from fetal to postnatal state in the patient with
a large defect is complex, with several vari-
ables that change during development. The
end result is the establishment of three genera]
states: high pulmonary flow-low resistance;
moderate flow-moderate resistance; and low
flow-high resistance. These states form the
continuum of groups 3, 4, and 5 (fig. 1). Car-
diac catheterization studies to date have docu-
mented only the end result, largely established
in the first few months of life and, with few ex-
ceptions, hemodynamically stable for 10 to 20
years. Critical experiments outlining the phases
leading to groups 3, 4, and 5, in the newborn
infant are lacking, but hopefully forthcoming.
Nevertheless, evidence is available to give a
reasonable answer to the question raised
earlier in this report: What determines where
a given patient lies on the curve of figure 1?
The patients in groups 3, 4, and 5 have in
common a large defect.4' 50 The essential ele-
ments that determine their separate postnatal
courses lie in the rate of involution of fetal
pulmonary arterial media,51' 52 the reactivity or
vasoconstrictive potential of these arteries,4'
t3 54 and the development of both right and
left ventricular ejectile force.5' All three fac-
tors are subject to biologic variation. Each in-
fluences the other two. Finally, each is sub-
ject to change until a stable hemodynamic
state is realized (i.e., entry into groups 3, 4,
or 5).
Vasoconstriction is essential to the theory of
ventricular septal defect evolution. Paul
Wood 4',' repeatedly emphasized its role,
particularly in the evolution of the Eisenmen-
ger complex. He believed that the vasocon-
striction leading to the Eisenmenger complex
 946
originates from a large left-to-right shunt that
begins early in the neonatal period and before
pulmonary arterial medial involution takes
place. Edwards has stressed a second ori-
gin of early vasoconstrictive responses. He
has argued that it is the common ejectile force
into pulmonary artery which is dominant in
preventing involution of fetal lung architec-
ture.51
The author believes that elevations of pres-
sure in the pulmonary venous-left atrial svs-
tem provide a third element that plays an im-
portant role in the neonatal physiologic
changes. Cardiac catheter and surgical data
have shown high left atrial pressures in group-
4 patients. The levels may not be far above
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
normal limits but they have been recorded
after the protective pulmonary vascular reac-
tion has been established. Pulmonary arteri-
olar wedge pressures were analyzed in 46 cases
from groups 1, 2, and 4 and compared to
right atrial mean pressures.* In 32 cases from
groups 1 and 2 (right ventricular systolic
pressure < 40 mm. Hg; average pulmonary-
to-systemic flow ratio 1.85), the pulmonary
wedge-right atrial difference was 3.8 -+- 2.2
mm. Hg. In 14 group-4 cases (right ventricu-
lar systolic pressure > 40 mm. Hg; average
pulmonary-to-systemic flow ratio, 1.95), this
difference was 9.6 2.2 mm. Hg (p < 0.001).
This observation confirms the work of Du-
Shane and Kirklin,58 who observed high left
atrial pressures in group-4 patients at open-
heart surgery.
A hypothetical relation between pulmonary
vascular medial involution and the forces op-
posing it is pictured vectorially in figure
23A-D. The normal pulmonary arterial media
involutes completely in anywhere from 6
weeks to 3 or rarely 6 months.5l The different
rates are schematically shown in figure 23A.
In the presence of a large defect vasoconstric-
tive forces oppose involution. They may slow
it (fig. 23B), stop it completely (fig. 23C), or,
going beyond, promote secondary medial by-
* The difference between the mean pulmonary ar-
tery wedge pressure and the mean right atrial pres-
sures was used to eliminpte effects of instrument base-
line variation,
BLOOMFIELD
pertrophy (fig. 23D). It is theoretically pos-
sible that equal vasoconstrictive stimuli will
slow involution in one patient and stop it in
another. Similarly, the concept of differences
in vascular reactivity explains why shunts of
equal magnitude occur in the presence of vast-
ly different vasoconstrictive forces in separate
patients.
The evolution of groups 3, 4, and 5, for
patients having equal-sized defects, is depicted
in figure 24. Vasospastic influence is visualized
to have two thresholds: the vasospasm neces-
sary to maintain pulmonary artery pressure
at systemic levels and the vasospasm necessary
to stop completely the medial involution of
fetal vessels. The early natural history of each
of the three groups is visualized at three theo-
retical periods that lead into the steady state.
Times, shunts, and vasospastic units are por-
0
'c
o
._
I_
C " :LA
-D0
0
' EF
B
g
A -J
_LA
0
" LA 4EF
>JEF
C +1
Figure 23
Schemnatic diagram of the forces affecting involution
of fetal pulmonary artery media. A. Portrayal of bio-
logic variation among involution rates in normal in-
fants. B. Portrayal of ejectile forces (EF) and pul-
rnonary venonis-left atrial vasoconlstrictive forces (LA)
from a VSD slowing but not stopping medial involu-
tion in a subject who would normally involute rapidly.
C and D. Portrayal of forces stopping involution (C)
or reversing it (D,).
Circulation, Volume XXIX, Jnei 1964
 VENTRICULAR SEPTAL DEFECT 947
to maintain systemic pressure in the pulmo-
nary artery but he only slows medial involu-
hi
10 tion. The group-3 patient, a hyporeactor, does
(n not maintain systemic pressure in the pulmo-
Cf) nary artery and slows medial involution the
least.
0U
The evolution of group 5 is as follows. The
W initial severe pulmonary vasospastic reaction
J0 reduces the early shunt (time B). The left
J D atrial element of pulmonary vasospasm is re-
duced. By time B, however, two changes have
LL
occurred: ejectile forces have increased above
neonatal levels; and pulmonary media has not
(1)9 only failed to involute but has begun to hyper-
trophy. The vasoconstrictive reaction is now
largely maintained by ejectile forces. At time
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

100 19
C, the Eisenmenger reaction is fixed. The
H
A A
shunt is small, left atrial component is nil, but
INVOLUTION ' THRESHOLD ____
the ejectile forces of systemic pressure against
L1 SY~STEMIC PR -*- --
SURE THRES'H OLD the persistent, thickened fetal media are suf-
ficient to maintain systemic pressure in the
W pulmonary artery without medial involution.
Fetal architecture51 is thus maintained
W cn

I
w:
Eil 1Lm Ii through a rational physiologic mechanism; not
because of a congenital malformation.59
The group-4 patient, who at time A was
able to maintain systemic pressure but only
slowed medial involution, progresses differ-
ently. At time B (second blocks), because
H10 there has been some medial involution and
L19
the thrust-volume capabilities of the ventricles
H5 have increased, the patient increases his shunt.
J-
Because of partial involution of the media, the
GROUP 5 4 3 5 4 3 5 4 3 thrust-volume vasospastic element is weak-
AGE A B C ened, but left atrial flooding is exaggerated.
Figure 24 Congestive failuire may occur with further
Schematic diagram illustrating the evolution of large
elevation of left atrial pressure. The left atrial-
defects, equal in size, into three distinct groups wtith pulmonary venous vasospastic element in-
separate natural histories. creases enough to maintain systemic pressure
in the lesser circulation and perhaps begins to
trayed schematically. The vasospastic influ- arrest medial involution. By time C (third
ence of the ejectile forces (pressure and vol- block) secondary medial hypertrophy has be-
ume) are diagramed as "EF" and the influence gun, increasing the thrust-volume vasospastic
of pulmonary venous-left atrial pressure as element. The shunt has diminished but re-
"LA." mains large. An element of left atrial pressure
In the first set of blocks (time A), all groups elevation remains, which we measure with a
have equal, large, left-to-right shunts. The cardiac catheter, and the sum of thrust and
hyperreactor (group 5) responds with suffi- left atrial influences is adequate to maintain
cient vasoconstriction to prevent medial invo- a steady hemodynamic state.
lution. The group-4 patient responds enough The group-3 patient, a hyporeactor, never
Circulation, Volume XXIX, June 1964
 948
develops sufficient vasospasm to prevent
medial involution or maintain systemic pres-
sure despite his large shunt. This patient is
most susceptible to early death from conges-
tive failure. As the media involutes, the shunt
becomes larger and either the left atrial-pul-
monary venous reaction increases sufficiently
to prevent further shunt increase, or the shunt
in this case is finally limited by defect size.
Once the relation between shunt, pulmonary
artery pressure, pulmonary vascular resistance,
and pulmonary venous pressure is established,
the children, with few exceptions, enter a
period of hemodynamic stability lasting one or
two decades. However, the critical question is:
Does pulmonary vascular pathology remain
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
as stable as serial hemodynamic studies
indicate?
Serial hemodynamic studies were performed
in 10 patients at the National Heart Hospital
(table 7). The ages of these patients varied
from 4 to 50 years at the time of their first
study and the interval between studies aver-
aged 8.6 years. Although this interval is longer
than preceding studies,'6-18 the findings are in
general agreement with them. Two cases
showed increased pulmonary vascular resist-
ance, one between age 5 and 12, the second
between ages 21 and 31 (cases DC-4 and DC-
10). These cases are discussed below. Two
older patients with small-to-moderate systolic
pressure gradients between the right ventricle
and the pulmonary artery showed increases in
these gradients (cases DC-2 and DC-5). One
case (DC-3) showed an apparent fall in pul-
monary vascular resistance and increased
shunt, although she was clinically unchanged.*
* Simple recording of serial cardiac catheterization
without reference to simultaneous clinical evaluation
can be misleading. Two examples (cases DC-3 and
DC-9, table 7) have already been discussed in the
text. Case DC-10, table 7, is a third case in point.
The patient was already clearly cyanotic on exertion
at the time of his first catheterization at age 5. Yet
calculated values inidicated a pulmonary-to-systemic
ratio of 3.7. This value combined with a calculated
resistance of 8.3 uInits plots as a misfit inl figure 1.
The pulmoinary flow certainly appears to be too higlh
for the calculated resistance. It is likely that the actual
pulmonary flow was less than calculated, and pul-
monary resistance was indeed higher than 8.3 units
9BLOOMFIELD
Hemodynamic data in the remaining five cases
remained constant.
The relation of pulmonary vascular resist-
ance to age found in our study is plotted in
figure 25 for 83 patients with pulmonary hy-
pertension. Groups 1 and 2 have been omitted,
since they would clutter the bottom of the
graph from ages 3 to 50. Arrows connect serial
catheterizations. No correlation of pulmonary
vascular resistance with age exists except
within group 5.t Eight of the 14 patients in
group 5 were known to be cyanotic from
infancy. The early onset of the Eisenmenger
complex in the majority of such patients is well
at first catheterization. The effect is to show a rise in
calculated pulmonary vascular resistance more appar-
ent than real.
Similar inconsistencies can be seen from other
studies, particularly where it is claimed that pulmo-
nary vascular resistance has increased within the pedi-
atric age groups. Case 26 of Lynfield et al.,17 case 31
of Lucas et al.,52 and cases 1, 6, 8, 11, 13, 14, and 17
of Auld et al.62 show pulmonary flow values that seem
excessive for a given resistance.
The methods for calculation and comparison of
pulmonary vascular resistances will also affect inter-
pretations. The author prefers evaluation of pulmo-
nary resistance, at least in children 3 years old and
over, in terms of "-units."-- "Units" is a simple term
relating pressure drop across the pulmonary bed, and
the flow through that bed. It is not clear that its
derived equivalent "dyne sec. cm. 5," related or not
to body surface area, provides any more information
concerning the pathology or resistance to flow of the
pulmonary vascular bed. When resistance values are
converted to units, many reported "increases" in pul-
monary vascular resistance 17, 52. 62 become insignifi-
cant (i.e., case 26,17 cases 32 and 34,52 and cases 2,
6, 8, 9, 13, and 14 62).
These examples from our work and from the
literature point out the difficulties, if not the folly, of
determining the natural history of ventricular septal
defect on the basis of serial cardiac catheterization
alone.
t The slope for the 83 points in figure 25 is +0.087
units per year of age with a correlation coefficient of
0.15. These figures are heavily weighted by the ]1.5
valuies from group 5. The slope from the 68 points in
giroups .3 anid 4 alonie is actually negative, 0.090,
wvith a correlation coefficienit of 0.29. rhe 15 group-5
patients hald a slope of +0.45 units per year of age
with a correlation coefficient of 0.63. For all calcutla-
tions, 35 units was considered maximal ptulmonary
resistance.
Czrculation7, Volun?e XXIX, June 1964
 VENTRICULAR SEPTAIL DEFECT 949

GROUP 3 4 5 * a
this documentation by serial cardiac catheter-
SYMBOL 0 0 . ization alone is incomplete and prompts the
3C
withholding of judgment.
1
The chief difficulty with figure 25, our 8- to
z
1
11 12-year follow-up studies, and the observa-
U) 1
1
tions of others 16-18 is that the follow-up peri-
1
z 1 ods have been too short. Referring back to ta-
20 i

a 11
1 ble 2 and figure 2, and comparing age distri-
Lu 1
1
'
bution of groups 4 and 5, we find that the late
I
i
1
11
peak of group-S patients is statistically sig-
i
i '
nificant (p <0.025). The paucity of group-4
e IC
patients after age 25 and the relative increase
0 0
9 0
0

&._
cr cr:.* -
0
0 ;---A*
0 0 0
in group-5 patients together suggest that the
0 0
0 0
. 0 late peak within group 5 is at least partially
0
0

0 0
0 0
0

0
0
20
o',
0

0-o--
25
-
-

OZI attributable to progressive pulmonary hyper-

1-- 1 1

5 10 15 20 25 30 40) 50 tension, the acquired Eisenmenger complex.

Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

AGE
Figure 25
Relation between pulmonary vascular resistance and
age in defect groups 3, 4, and 5. Dotted lines repre-
sent results of serial catheterizations. (Cases DC 4, 6,
7, 10, table 7.)
known.4 59 Only three gave a history suggest-
ing possible progression to group 5 from
groups 3 or 4.
Documentation of progressive increase in
pulmonary vascular resistance by serial cardi-
ac catheterization is not common. From 10
cases, ages 3 to 133E, reported in the litera-
ture 17, 52, 57 prior to the report of Auld et al.,62
a pattern emerges defining those who are vul-
nerable: all were high pressure-high resistance
group-4 defects at first cardiac catheterization.
None had a calculated pulmonary resistance
less than 7.4 units. All had mean pulmonary
artery pressures close or equal to systemic
pressures. The data of Auld et al.62 contrast
with these findings. On the basis of serial car-
diac catheterizations they reported increases
in pulmonary vascular resistances in 11 of
12 cases of isolated defect. Two of their pa-
tients (cases 7 and 10) could be examples of
low pressure-low resistance changing to high
pressure-high resistance levels, but in view
of the absence of reports of any similar cases,
* Auld et al.62 are incorrect when they state '<A
proportion of the cases reported by Lucas et al.5'
showed changes similar ." for every case reported
by Lucas et al., in which pulmonary vascular resist-
Ci.rculation, Volume XXIX, June 1964
- YEARS
Therefore, although hemodynamic stability is
the rule during the first 20 years of life, it is
probable that some permanent vascular
changes are occurring that cannot be measured
by the technics of cardiac catheterization. This
probability is consistent with the work of Ed-
wards 51 and recent studies suggesting slow
resolution of elevated pulmonary vascular re-
sistances in patients after surgical closure of
the defect.52 66 These changes begin to take
their toll after age 20.
The same may be said for patients in group
3, although this group is too rare for thorough
documentation. Case DC-4, table 7, is illustra-
tive. Although pulmonary vascular resistance
increased from 1.1 to 4.7 units between ages
21 and 31, the change appeared to be largely
vasoconstrictive, for the patient's pulmonary
vasculature remained acutely sensitive to the
administration of oxygen. Oxygen doubled
pulmonary flow and halved pulmonary resist-
ance rose, fell into the vulnerable high pressure-high
resistance pattern. It should also be noted that the
average calculated resistance change (units) in the
seven cases of simple defect (cases 2, 6, 8, 9, 13, 14,
and 17) of Auld et al. who were vulnerable (i.e.,
high pressure-high resistance) at the time of their
first cardiac catheterization, is -0.7 units. In other
words. the data of Auld et al.02 limited to isolated
defect and translated into the terms of our paper
oddly suggest that pulmonary vascular resistance in-
creased only in those with initially low values, and
. . remained constant in that group of patients with
initially high values.
 950
ance. Vhatever changes took place in this
group-3 patient, they were gradual, mild, and
not fixed, even at age 31. The course of group-
3 patients appears similar to that of atrial sep-
tal defect, except that the prognosis is less
favorable because of the chronic volume over-
load on the left ventricle.
An additional influence upon the progres-
sion of pulmonary vascular disease that must
be considered is evidenced by the work of
Davies.63 He first noted the chest deformity
frequently seen in congenital heart disease,
and particularly ventricular septal defect, with
hyperkinetic pulmonary hypertension. Con-
tinuing this work, Davies et al.4 found in-
creased lung stiffness in such cases and docu-
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
mented that this particular complication was
associated with a high incidence of lung in-
fection in childhood. Some of the cases de-
scribed showed gross disturbances of lung
volume, ventilation, and diffusion. These find-
ings, recently confirmed by Woolf,65 can only
aggravate pulmonary vascular disease by im-
posing an element of cor pulmonale on the al-
ready overloaded heart.
Summary
It is now possible to synthesize the natural
history of ventricular septal defect. Size is
fundamental, and stemming from it, the small-
er defects (groups 1 and 2) separate from the
larger (groups 3, 4, and 5) immediately at
birth. Apparent size may differ from real size,
for defects may become partially closed by
tricuspid valve leaflets, or be gradually ob-
structed by an enlarging aneurysm of the sinus
of Valsalva. One instance has been observed in
wsrhich a muscular defect was obstructed by a
large papillary muscle. The reverse may also
occur, i.e., a partially obstructed defect may
enlarge as a blocking tricuspid leaflet detaches
or tears.
Natural History of Small Defects
Groups 1 and 2 can be discussed together.
Growth and development are normal. Early
problems such as congestive failure and lung
infections are not common. Few have symp-
toms attributable to the defect. Approximately
25 per cent of defects present in these patients
BLOOMFIELD
at the age of school entry vill close spon-
taneously during their lifetime. The major risk
for these groups is bacterial endocarditis,
which has a peak incidence in the third and
fourth decades. The risk has been shown to
be approximately 30 per cent for a 70-year
lifetime and should be considerably lower with
judicious penicillin prophylaxis. The lifetime
mortal risk of bacterial endocarditis in these
groups is less than a per cent.
Group-2 patients will gradually approach
group-1 status clinically in the third and fourth
decades, if not sooner. As the aortic root
gradually enlarges with age, the proportion
of left ventricular output shunting falls. The
increased chest diameter obscures the left
ventricular impulse, ventricular filling mur-
mur, and the Roger murmur itself. Diagnosis
is more difficult over age 40. These patients
seldom die young and they do not develop
pulmonary hypertension. They visit our adult
heart clinics infrequently because of the fol-
lowing factors: (1) they feel well (it was
shown that six of seven patients followed-up
after 24 years would have been missed in a
survey of adult defects) (2) spontaneous clo-
sure (3) increased diagnostic difficulties.
A complication of aging for the patient with
defect, open or closed, has been shown to be
increased susceptibility to other left ventricu-
lar diseases. It is believed that the defective
septum, open, spontaneously closed, and per-
haps surgically closed, cannot make its normal
contribution to ventricular ejection. In tle
presence of left ventricular disease, such as
ischemic heart disease or hypertension, this
weakness may initiate early congestive heart
failure. Otherwise, the defect may be of trivial
import throughout life.
The anticipated life expectancy for the 95
per cent of patients in groups 1 and 2 surviving
bacterial endocarditis is perhap-s 65 years. With
a possible 5-per cent mortality from bacterial
endocarditis at age 30, the over-all life ex-
pectancy is 63 years.
The surgical approach to groups 1 and
2 is mainly prophylactic against bacterial en-
docarditis. It is too early to predict that there
will be a decreased incidence of susceptibil-
ity to left ventricular failure, although this is
Circulation, Volume XXIX, June 1964
 VENTRICULAR SEPTAL DEFECT
possible. Surgical treatment for these patients
should be placed in proper perspective. Since
the life expectancy of these groups untreated
will average 63 years, it is sobering to realize
that nearly a century will pass before the
patient-years of those whose lives have been
extended by surgery will exceed the patient-
years already lost to it. For this reason, when
group-i and group-2 patients are treated surgi-
cally, the established risk must be below 1 per
cent. Gains to be made in these patients are
very small, and losses, when they occur, are
large.
Natural History of Large Defects
The outlook and prognosis for pulmonary
hypertensive ventricular septal defects of our
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
study confirm the work of others.9' 13, 16-18, 57
The childhood course of groups 3 and 4,
commonly complicated by congestive failure
in the first year of life, a high incidence of
pneumonia and other respiratory infections,
moderate effort intolerance or fatigue, and
retarded growth, has been well detailed by
others.9' 17, 18 Group-5 patients are similar.
They may have more effort intolerance and
cyanosis, but less congestive failure and respir-
atory symptomatology.4 63 64 While groups 3,
4, and 5 are separate physiologic variants in
the first two or three decades of life, they tend
to fuse toward a single syndrome after age 30,
which is characterized chiefly by gradually
increasing effort dyspnea and varying degrees
of bidirectional shunting. Pathologic stud-
ies 13, 51 have concentrated intently upon aortic
dextroposition and details of pulmonary vascu-
lature. The group entities of table 1 cannot
be sorted from available pathologic data, par-
ticularly in adults.
The mean survivorship of large defects in
Table 17
Comparison of Excess Cardiac and Pulmonary Stress in Pulmonary Hypertensive
Ventricular Septal Defect (Groups 3, 4, and 5)
Left ventricle
excess load
Group Volume Pressure
3 +++ N ++
4 N
5 N N N
N, normal; +, +-+, +++, mild, moderate, and
Circulation, Volume XXIX, June 1964
951
the series of Selzer and Laqueur 13 is 31 years
and that of Wood4 33 years. Their cases were
collected from a wide variety of sources, some
from the autopsy tables of the nineteenth cen-
tury,13 and some from the operating tables
of the twentieth.4 This was necessary because
natural deaths among these patients surviving
infancy during a period of continuous observa-
tion by a single physician is uncommon. Only
a single natural death was reported in the 7-
year experience of Fyler et al.9 from Chil-
dren's Medical Center, Boston, among children
surviving infancy. In the series of table 12,
four out of five patients with large defects sur-
vived their fortieth birthday. In 1963 it is
probable that mean survivorship in large de-
fects surviving infancy is 35 years.
There may be differences in survivorship
among group-S patients, depending on wheth-
er the Eisenmenger complex was acquired
early or late. Patients from groups 3 and 4
who acquired it after 20 years might be more
prone to early congestive failure. This is
because the over-all cardiac work, illustrat-
ed schematically in table 17, for group-5
patients is less than that for groups 3 and 4.
Table 17 illustrates the excess biventricular
and pulmonary stresses to which groups 3 and
4 are subject. Survivorship among the five
group-4 patients of our 8- to 12-year follow-up
study was poor. However, only two of the
four deaths (at age 8 from bacterial endo-
carditis, and at age 13 from heart failure and
pneumonia) fit within the definition of this
natural history. Dr. Wood recorded only two
natural deaths (at ages 25 and 32) among 44
group-3 and group-4 patients he had followed
for periods up to 10 years,"' and one of these
was due to cancer.
Right ventricle
excess load Lung
Volume Pressure stiffness 64
+-++ ++
+
N
severe changes, respectively.
 952 BLOOMFIELD

Half the patients reported died of conges-

tive failure, frequently complicated by pneu-
inonia. Another quarter died of bacterial endo-
carditis. Bacterial endocarditis is less commoii
in hypertensive than smaller defects but it is
by no means rare. When it occurs, it is a grave

bt'' *^,0 l0 O O

ccO

rq 00
I'_ tO tm~1m~i
LZ C]
complication. Hemoptysis is a terminal event
in 25 per cent, often complicating congestive
failure, pneumonia, or endocarditis. The re-
mainder of patients with large defects die of
unusual events such as ruptured aneurysms of
the sinus of Valsalva or aneurysms of the pul-
i-
monary artery, cerebral abscess, or cerebral 0 000 000
0II
C
c
C 3 C

thrombiosis.4 1'
Surgical Considerations in Large Defects
Surgical closure is clearly indicated for
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

groups 3 and 4 and contraindicated for group

5 at the present time. Two schools of thought
influence surgical timing. One advocates pa-
tience,1619 by stressing the hemodynamic sta-
bility of these lesions. The other is concerned
lest patients operable at age 1 become inoper-
able before ages 5 or 10 by virtue of irre- cc
versible pulmonary vascular change.6 67, 68
Table 18 has been constructed to view pul-
M-
00 be
a)~ iCEEc ~~~~~~~~~~~~~~~c

monary hypertensive and surgical risks from FS.

an actuarial approach. The figures of table 18
show the theoretical cumulative survival of U v.'R-
100 patients with large defects operated upon I~~~~~~~~~~~~~~~~~~c

either early at age 1 or late at age 7. The as-

sumptions of table 18 include the following. a) -~~~~~~~~~~~~~~~~~~~c
1. Twenty per cent of pulmonary hypertensive
patients become inoperable (a high figure) Z1~~~~~~~~~~~~~~~a

between ages 1 and 7. The mean survival for 10 u :)

these patients is 35 years. 2. The life expec- PT.(
tancy of surgically treated large defect is 65 U~~~~~~~~~~~~~~~C ct:
years.* 3. Operative moortality at age 1 is 20 CirulaionVoumeXXI, jne 96
per cent, and at age 7 is 5 per cent. 4. No
deaths occur in nonoperated patients between
ages 1 and 7. Table 18 also excludes medical
and surgical morbidity. C-,' c 1

It will be seen from table 18 that the cumu-

lative survival to age 65 in patient-years
*
The survivorship of surgically closed large ven-
tricular septal defects remains to be answered. In
view of cases N-18, 19, 20, 21, and 23, table 12, and
the cardiomyopathic changes observed in older pa-
tients, a mean survivorship of 65 years is a maximal
figure.
 VENTRICULAR SEPTAL DEFECT
vill be longer in the group of children not
subjected to the higher risk of early surgery,
even though as many as 20 per cent become
inoperable by age 7. The excess mortality at
age 65 in the early operation group is 448
patient-years per 100 patients per lifetime.
Since the risk of any piatient becoming in-
operable between ages 1 and 7 is actually
closer to 5 per cent, early surgery with ele-
vated risk is rarely indicated. The author
agrees with Nadas et al. that "fear of progres-
sion of pulmonary vascular disease should not
be a deciding factor in timing surgery."'9
Complications of Ventricular Septal Defect
in Relation to Natural History
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
It was our purpose to avoid complicating
lesions, but two deserve mention. Aortic in-
sufficiency is confirmed as a grave complica-
tion regardless of defect severity.'3 57 In-
fundibular pulmonary stenosis was observed
to increase abruptly in one patient after age
40. The patient previously had mild pulmonary
valvular stenosis. Although this has been re-
ported previously in infants,9' 17 it has not been
observed in adults.
Acknowledgment
This investigation was begun during the tenure
of a l-year Special Fellowship from the National
Heart Institute, United States Public Health Service,
and was completed during the third year of tenure
of an Established Investigatorship of the American
Heart Association. It was supported in part by grants
from the Heart Societies of Cleveland, Portage Coun-
ty, and Western Counties of Ohio.
I am particularly indebted to two colleagues: Dr.
Hywel Davies, for the many hours of fruitful discus-
sion and helpful suggestions during the course of the
work, and Dr. Jerome Liebman, who painstakingly
reviewed the manuscript, contributing to it in count-
less ways.
References
1. COOLEY, D. A.: Current status of surgical treat-
ment of VSD. Dis. Chest 35: 651, 1959.
2. ROGER, H.: Recherches cliniques sur la com-
munication congenitale des deux coeurs, par
inocclusion du septum interventriculaire. Bull.
Acad. Med. Paris, 2nd ser., 8: 1074, 1879.
3. WOOD, P.: Diseases of the Heart and Circulation.
London, Eyre & Spottiswoode, 1959.
4. WOOD, P.: The Eisenmenger syndrome. Brit.
M. J. 2: 701, 755, 1958.
Circulation, Volume XXIX, June 1964
953
5. Quoted by Documenta Geigy, Ed. 5, New York,
1959, p. 249, from work of Boothby, W. M.,
Berkson, J., and Dunn, H. L.: Am. J. Physiol.
116: 468, 1936; and Dubois, D., and Dubois,
E. F.: Arch. Int. Med. 17: 863, 1916.
6. Ibid. Page 257, quoted by Stuart, H. C., and
Stevenson, S. S.: Physical growth and develop-
ment. In Nelson, W. E.: Mitchell-Nelson Text-
book of Pediatrics. Philadelphia, W. B.
Saunders Company, 1950.
7. WEBER, F.: Can the clinical manifestations of
congenital heart disease disappear with the
general growth and development of the pa-
tient? Brit. J. Child. Dis. 15: 113, 1918.
8. WOOD, P., MAGIDSON, 0., AND WILSON, P. A. 0.:
Ventricular septal defect with a note on
acyanotic Fallot's tetralogy. Brit. Heart J. 16:
387, 1954.
9. FYLER, D. C., RUDOLPH, A. M., WITTENBORG,
M. H., AND NADAS, A. S.: Ventricular septal
defect in infants and children: A correlation of
clinical, physiologic, and autopsy data. Circu-
lation 18: 833, 1958.
10. SELZER, A.: Defect of the ventricular septum:
Summary of the twelve cases and review of
the literature. Arch. Int. Med. 84: 798, 1949.
11. TERPLAN, K., AND SANES, S.: The incidence of
congenital heart lesions in infancy: A com-
parative statistical study based on post-mortem
examinations. J. Tech. Methods 15: 86, 1936.
12. KEITH, J. D., ROWE, R. D., AND VLAD, P.: Heart
Disease in Infancy and Childhood. New York,
Macmillan Company, 1958, p. 213.
13. SELZER, A., AND LAQUEUR, G. L.: The Eisen-
menger complex and its relation to the un-
complicated defect of the ventricular septum:
Review of thirty-five autopsied cases of Eisen-
menger's complex, including two new cases.
Arch. Int. Med. 87: 218, 1951.
14. SAPHIR, 0.: Gross examination of the heart. In-
jection of coronary arteries. Weight and
measurements of heart. In Pathology of the
Heart, S. E. Gould, Ed. Ed. 2. Springfield,
Illinois, Charles C Thomas, Publisher, 1960,
p. 1064.
15. DE LA CRUZ, M. V., ANSELMI, G., ROMERO, A.,
AND MONROY, G.: A qualitative and quantita-
tive study of the ventricles and great vessels
of normal children. Am. Heart J. 60: 675,
1960.
16. DOWNING, D. F.: Recatheterization of the right
heart in ventricular septal defect. Am. Heart
J. 57: 669, 1959.
17. LYNFIELD, J., GASUL, B. M., ARCILLA, R., AND
LUAN, L. L.: The natural history of ventricular
septal defects in infancy and childhood. Am. J.
Med. 30: 357, 1961.
18. STANTON, R. E., AND FYLER, D. C.: The natural
history of pulmonary hypertension in children
 954
with ventricular septal defects assessed by
serial right-heart catheterizatioin. Pediatrics 27:
621, 1961.
19. NADAS, A. S., RUDOLPH, A. M., AND GROSS, R. E.:
Editorial: Pulmonary arterial hypertension in
congenital heart disease. Circulation 22: 1041,
1960.
20. ESPINO-VELA, J., AND MATA, L. A.: Eisen-
menger's complex: A clinical and pathological
study of four cases. Am. Heart J. 51: 284,
1956.
2 1. DE CARVALHO AZEVEDO, A., NEY TOLEDO, A.,
DE CARVALHO, A. A., ZANIOLO, WV., DOHMANN,
H., AND ROUBACH, R.: Ventricular septal de-
fect: An example of its relative diminution.
Acta Cardiol. 13: 513, 1958.
22. EVANS, J. R., ROWE, R. D., AND KEITH, J. D.:
Spontaneous closure of ventricular septal de-
fects. Circulation 22: 1044, 1960.
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
23. MAJKA, M., RYAN, J., AND BONDY, D. C.: Spon-
taneous repair of a ventricular septal defect.
Canad. M. A. J. 82: 317, 1960.
24. BLUMENTHAL, S., GRIFFITHS, S. P., AND MORGAN,
B. C.: Bacterial endocarditis in children with
heart disease. Pediatrics 26: 993, 1960.
25. HARNED, H. S., JR., AND PETERS, R. M.: Spon-
taneous closing of ventricular septal defects:
Two case reports. Circulation 22: 760, 1960.
26. NADAS, A. S., SCOTT, L. P., HAUCK, A. J., AND
RUDOLPH, A. M.: Spontaneous functional clos-
ing of ventricular septal defects. Nev England
J. Med. 264: 309, 1961.
27. AGUSTSSON, M. H., GASUL, B. M., ARCILLA, R. A.,
BICOFF, J. P., AND MONCADA, R.: Spontaneous
closure of ventricular septal defect in eight
children demonstrated by serial cardiac cath-
eterizations and by angiocardiography. Ab-
tract, Circulation 24: 874, 1961.
28. BLOOMFIELD, D. K.: Spontaneous closure of
ventricular septal defect: Clinical and patho-
logical correlations. Abstract, Circulation 24:
890, 1961.
29. ROBERTS, WV. C., MoRRow, A. G., MASON, D. T..
AND BRAUNWALD, E.: Spontaneous closure of
ventricular septal defect: Anatomic proof in
an adult with tricuspid atresia. Circulation 27:
90, 1963.
30. EDWARDS, J. E., PARKER, R. L., BURCHELL, H. B.,
WOOD, E. H., AND BULBANIAN, A. H.: An Atlas
of Congenital Anomalies of the Heart and
Great Vessels. Springfield, Illinois, Charles C
Thomas, Publisher, 1953, p. 67.
31. CANNELL, D. E.: Congenital aneurysm of the in-
traventricular septum: Report of two cases.
Am. J. Path. 6: 477, 1930.
32. LEv, M., AND SAPHIR, 0.: Congenital aneurysm
of the membranous septum. Arch. Path. 25:
819, 1938.
33. ROGERS, H. MN., EVANS. I. C_. AND DOMEIER,
BLOOMFIELD
L. H.: Congenital aneurysm of the mem-
branous portion of the, ventricular septum:
Report of two cases. Am. Heart J. 43: 781,
1952.
34. CLARK, R. J., AND WHITE, P. D.: Congenital
aneurysmal defect of the membranous portion
of the ventricular septum associated with heart
block, ventricular flutter, Adams-Stokes syn-
drome and death. Circulation 5: 725, 1952.
35. HOFFMAN, J. I. E., RUDOLPH, A. M., NADAS, A. S.,
AND GROSS, R. E.: Pulmonic stenosis, ventricu-
lar septal defect, and right ventricular pressure
above systemic level. Circulation 22: 405,
1960.
36. GRIFFIN, G. D. J., AND ESSEX, H. E.: Experi-
mental production of interventricular septal
defects. Surg. Gynec. & Obst. 92: 325, 1951.
37. BENVENUTO, R., AND LEWIS, F. J.: Gradual
closure of interventricular septal defects. J.
Thoracic Surg. 37: 673, 1959.
38.
RAUH, L. WV.: The incidence of organic heart
disease in school children. Am. Heart J. 18:
705, 1939.
39. ROBINSON, S. J., AGGELER, D. M., AND DANILOFF,
G. T.: Heart disease in San Francisco school
children. J. Pediatrics 33: 49, 1948.
40. GARDINER, J. H., AND KEITH, J. D.: Prevalence of
heart disease in Toronto children: 1948-1949
Cardiac Registry. Pediatrics 7: 713, 1951.
41. MACMAHON, B., MCKEOWVN, T., AND RECORD.
R. G.: The incidence and life expectancy of
children with congenital heart disease. Brit.
Heart J. 15: 121, 1953.
42. FONTANA, R. S., AND EDWARDS, J. E.: Congenital
Cardiac Disease. A Review of 357 Cases
Studied Pathologically. Philadelphia, W. B.
Saunders Company, 1962.
43. KAPLAN, S., DAOUD, G. I., GLASS, I. H., SHEM-
TOB, A., AND MCGUIE, J.: Natural course of
ventricular septal defect. Abstract, Circulation
24: 968, 1961.
44. LAUBRY, C., ROUTIER, D., AND DE BALSAC, R.:
Grosse pulmonaire: Petit aorte; affection con-
genitale. Bull. et mem. Soc. med hop. Paris 56:
847, 1941.
45. GELFMAN, R., AND LEVINE, S. A.: The incidence
of acute and subacute bacterial endocarditis
in congenital heart disease. Am. J. NI. Sc. 204:
324, 1942.
46. BLOOMFIELD, D. K.: Association of left ventricu-
lar myopathy with minor congenital defects of
the ventricular septum. Abstract, Circulation
24: 889, 1961.
47. HOLT, M., AND ORAM, S.: Familial heart disease
with skeletal malformations. Brit. Heart J. 22:
236, 1960.
48. BROCK, R.: Quoted in Documenta Geigy, The
WVorld of Cardiology, New York, 1962.
Circelation, Vo/lnme XXIX, June 1964
 VENTRICULAR SEPTAL DEFECT
49. RUDOLPH, A. M., AND NADAS, A. S.: The pulmo-
nary cIrculation and congenital heart disease:
Considerations of the role of the pulmonary
circulation in certain systemic-pulmonary com-
munications. New England J. Med. 267: 968
and 1022, 1962.
50. SWAN, H. J. C., SAVARD, R. M., AND KIRKLIN,
J. W.: Relation of size of ventricular septal
defects to circulatory dynamics. Abstract, J.
Clin. Invest. 37: 934, 1958.
51. EDWARDS, J. E.: Functional pathology of the
pulmonary vascular tree in congenital cardiac
disease. C rculation 15: 164, 1957.
52. LUCAS, R. V., JR., ADAMS, P., JR., ANDERSON,
R. C., MEYNE, N. G., LILLEHEI, C. W., AND
VARCO, R. L.: The natural history of isolated
ventricular septal defect: A serial physiologic
study. Circulation 24: 1372, 1961.
53. SHEPHERD, J. T.: The pulmonary circulation in
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017
the presence of interatrial, interventricular
and interarterial communications. In Pulmo-
nary Circulation, W. Adams and I. Veith, Eds.
New York, Grune & Stratton, 1959, p. 204.
54. DAWES, G. S.: In Pulmonary Circulation, W.
Adams and I. Veith, Eds. New York, Grune &
Stratton, 1959, p. 300.
55. XVOOD, P.: Pulmonary hypertension. Brit. M. Bulll.
8: 348, 1952.
56. WOOD, P.: Pulmonary hypertension with special
reference to the vaso-constrictive factor. Brit.
Heart J. 20: 557, 1958.
57. BROTMACHER, L., AND CAMPBELL, M.: The
natural history of ventricular septal defect.
Brit. Heart J. 20: 97, 1958.
58. DUSHANE, J. W., AND KIRKLIN, J. W.: Selection
for surgery of patients with ventricular septal
Circulation, Volume XXIX, June 1964
955
defect and pulmonary hypertension. Circula-
tion 21: 13, 1960.
59. KOHOUT, F. W., SILBER, E. N., SCHLICHTER, J. G.,
AND KATZ, L. N.: The dynamics of the Eisen-
menger complex. IL. Am. Heart J. 50: 337,
1955.
60. BURCHELL, H. B.: Studies in pulmonary hyper-
tension in congenital heart disease. Brit. Heart
J. 21: 255, 1959.
61. WOOD, P.: Posthumous papers.
62. AULD, P. A. M., JOHNSON, A. L., GIBBONS, J. E.,
AND MCGREGOR, M.: Changes in pulmonary
vascular resistance in infants and children with
left-to-right intracardiac shunts. Circulation 27:
257, 1963.
63. DAVIES, H.: Chest deformities in congenital heart
disease. Brit. J. Dis. Chest 53: 151, 1959.
64. DAVIES, H., WILLIAMS, J., AND WOOD, P.: Lung
stiffness in states of ahnormal pulmonary
blood flow and pressure. Brit. Heart J. 24: 129,
1962.
65. WOOLF, C. R.: Pulmonary function in adults
with intracardiac septal defect. C rculation 27:
261, 1963.
66. BRAUNWALD, N. S., BRAUNWALD, E., AND MOR-
ROW, A. G.: The effects of surgical abolition
of left-to-right shunts on the pulmonary vascu-
lar dynamics of patients with pulmonary hyper-
tension. Circulation 26: 1270, 1962.
67. KAY, E. B., ZIMMERMAN, H. A., AND CRoss, F. S.:
Considerations in the surgical treatment of
ventricular septal defects. J. Thoracic Surg. 30:
452, 1955.
68. ADAMS, P., ANDERSON, R. C., ALLEN, P., AND
LILLEHEI, C. WV.: Physiologic changes with
age in ventricular septal defect. Abstract, Cir-
culation 16: 8,7, 1957.
 The Natural History of Ventricular Septal Defect in Patients Surviving Infancy
DANIEL K. BLOOMFIELD

Circulation. 1964;29:914-955
Downloaded from http://circ.ahajournals.org/ by guest on January 6, 2017

doi: 10.1161/01.CIR.29.6.914
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright 1964 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/content/29/6/914

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles

originally published in Circulation can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column of
the Web page under Services. Further information about this process is available in the Permissions
and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Circulation is online at:

http://circ.ahajournals.org//subscriptions/