Download Cardiac Volume in Normal Children and Adolescents

Cardiac Volume in Normal Children
and Adolescents
Its Application to Patients with Rheumatic
Mitral Insufficiency
By QUANG X. NGHIEM, M.D., MELVYN H. SCHREIBER, M.D.,
AND
LEONARD C. HARRIS, M.D., M.R.C.P.
SUMMARY
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Cardiac volume was determined by a simplified Rohrer-Kahlstorf method in 305
healthy children and adolescents. The ages ranged from birth to 19 years. Height and
weight were comparable with normal growth standards. Race, sex, age, height, weight,
and body surface area were studied for their value in predicting cardiac volume. Cardiac
volume was different in males and females (P < 0.005) and was more closely predicted
by weight than by body surface area. In this study body weight has been used as a
single predictor of cardiac volume for both sexes. Normal values are presented and the
limitations of the technique discussed.
Serial cardiac volume plotted against weight and cardiothoracic ratio plotted against
time were correlated with the clinical course in 27 children with pure rheumatic mitral
insufficiency. Cardiac volume was found to be superior to the cardiothoracic ratio in
reflecting the severity and prognosis of mitral insufficiency. After the cardiac volume
had reached a level of 1,100 to 1,300 ml, the course was progressively downhill in 10
adolescents in the absence of evidence of rheumatic activity in most cases.
That the course of rheumatic mitral insufficiency is essentially volume-dependent is a
phenomenon of great interest which may be used to assess the severity of mitral regurgitation and its future course.
ADDITIONAL INDEXING WORDS:
Teleoroentgenograms
Cardiothoracic ratio
Cardiac enlargement
cardiac volume.3-5 While normal standards
have been determined for adults,612 there
are no satisfactory pediatric data13 except for
infants.'4 Thus, whether the cardiac volume
correlates best with body surface area or
other parameters of growth in normal children has not been made clear.'3 The usefulness of cardiac volume has even been questioned.'5 Nevertheless, when confronted with
a patient with cardiac enlargement, both the
clinician and the radiologist require a reasonably accurate and reproducible method of
determining heart size.
One purpose of this study is to establish
standards for cardiac volume in healt-hy subjects from birth through adolescence when
the growth spurt ceases. Although this is not
ROENTGENOGRAPHIC projection techniques for determination of the volume
of the heart were proposed as early as 1916
by Rohrer' and in 1932 by Kahlstorf.2 Subsequent methods based on their work have
been used clinically in Europe. Recently, there
has been a renewal of interest in studies of
From the Departments of Pediatrics and Radiology,
University of Texas Medical Branch, Galveston,
Texas.
Work was supported in part by Grant 5 Ti HE555703 from the National Institutes of Health, U. S.
Public Health Service, and by grants from the Texas
and Bay Area Heart Associations. Computation facilities provided by Grant 5-P07 FR-00024 from the National Institutes of Health were used in analyses
of data.
Circulation, Volume XXXV, March 1967
509
NGHIEM ET AL.
510
a longitudinal study, our samples of different
age groups are comparable with normal
growth data compiled by others.'6 '1 Relationships have been studied between cardiac
volume, on the one hand, and body surface
area, weight, height, age, race, and sex on the
other. A second aim is to test the superiority
of cardiac volume over cardiothoracic ratio
as an index of the severity and future course
of rheumatic mitral insufficiency. Our experience indicates that not only the ultimate outcome but also the clinical course of
rheumatic insufficiency are related to the
volume of the heart.'8
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Methods
and
hundred
fifty-one sets of teleoroentFour
genograms of the chest in posteroanterior and
lateral projections, coded as normal at the University of Texas Medical Branch, were reviewed
to ensure that there was no radiographic evidence
of cardiovascular or respiratory disease. All the
films were exposed with the patient erect except
those of small infants in whom anteroposterior
supine films were made. The cardiac volume was
estimated in each case by a simplified method.'9
Detailed clinical records of each subject were
reviewed. Those with clinical evidence of cardiopulmonary disease and those who were obese,
pregnant, malnourished, or dehydrated, or who
had abnormal blood counts or urinalyses were
excluded. Subjects whose height or weight was
not compatible with growth data of Stuartl6 or
of Falknerl7 were excluded from the study also.
Three hundred and five subjects were studied.
They were attending outpatient clinics or were
hospitalized while awaiting elective surgery (for
example, repair of hernias and burn scars).
There were 206 Caucasians (including Latin
Americans) and 99 Negroes; 158 were males
and 147 females. Ages ranged from birth to 18
years and 9 months (fig. I). Body surface area
ranged from 0.18 to 1.87 M2, and body weight
from 6.2 to 167 lb. Forty-six subjects with
height or weight slightly above or below the
fifth percentile were separately analyzed. Since
Age Distribution of 305 Normal Subjects Studied
15-
Female Male
HI
I10
En
U
V
Age in months or years
Figure 1
The age distribution of the study sample is comparable with the body weight distribution (see
figs. 2 and 3).
Circulation, Volumne XXXV, Marcb 1967
CARDIAC VOLUME
511
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
the prediction of cardiac volume was not affected by extremes of height, relatively small or
tall children were included in order to make up
a sample representing a healthy population at
large. Body surface area was computed from the
DuBois and DuBois formula,20 and all data
were statistically analyzed.
In 27 cases of rheumatic mitral insufficiency
previously described,18 the clinical course was
correlated with serially determined cardiothoracic ratios and cardiac volumes. During hospitalization for active carditis or cardiac failure,
chest roentgenograms were performed monthly
or even weekly if cardiac volume was changing
rapidly. For outpatients, cardiac volume determinations were usually made at 6 to 12-month
intervals. In six of the seven fatal cases, terminal
events were documented and in two autopsy was
performed. For the total group, the mean duration of follow-up from the initial attack of
rheumatic fever was 5 years and 3 months;
the median was 4 years and 2 months, and the
range was 13 months (patient deceased) to 12
years and 3 months.
Jones' criteria (revised)21 were used to define
recurrent attacks of rheumatic fever which were
presumed to be associated with carditis. Congestive heart failure per se or further enlargement of the heart without pericarditis or involvement of structures previously considered normal
(for example, tricuspid valve) were not regarded
as indicating a new attack of carditis without
other evidence. Left atrial biopsies, mitral valve
tissue in two patients, and autopsy material in
two others were reviewed by a pathologist who
had no knowledge of previously recorded diagnoses. The pathological criteria for rheumatic
activity were those of Gross and Ehrlich.22
One hundred and ninety-eight determinations
of cardiac volume were made and plotted
against patients' weights to determine the trends
of cardiac volume changes. In 179 posteroanterior chest roentgenograms, cardiothoracic ratio
and cardiac volume per square meter of body
surface area (cardiac volume index) were compared as indicators of cardiac enlargement.
The simplified method of estimating cardiac
volume was based upon the principle of approximating the heart to an ellipsoid, using a constant magnification factor for the 6-foot distance
between target and film (183 cm) .19
Cardiac volume =
MxLxWxD
= 0.44 x L x W x D
where M
or
-X
the correction factor for magnifica-
tion
1.19
Circulation,
Volume XXXV,
March
1967
L = long diameter of the heart, from the junction of the superior vena cava with the
right atrium, to the cardiac apex.
W=width of the heart: sum of the distances
at right angles to the long diameter of the
heart, to the cardiophrenic angle on the
right, and to the cardiovascular junction
on the left.
D = depth of the heart from the most posterior
aspect of the heart shadow to the posterior
border of the sternum.
Reproducibility of the measurements of cardiac volume was evaluated by comparing the
results of the observer reading the total sample
(M.H.S.) with those of the other two authors.
Delineation of cardiac axes was left to individual discretion. Ninety-three sets of chest roentgenograms selected from all age groups from the
total of 305 subjects were independently examined. The value of a predictor of cardiac
volume, consistent rater bias, and random
error of measurements were studied.
The cardiothoracic ratio was defined as the
sum of the maximal distances at right angles
from the midthoracic plane (midline of the spine)
to the farthest right and left cardiac contours,
divided by the maximal internal diameter of the
chest at the level of the diaphragm.23 The transverse diameter of the chest at the level of the
right diaphragmatic dome could not be used
because the dome itself was included in the
cardiac shadow with severe cardiac enlargement.
Results
In 261 cases in which body surface area
was known, correlation coefficients of various parameters with cardiac volume are given
in table 1. No difference was found between
cardiac volume in whites and Negroes. Multiple regression analysis24 showed that age
did not contribute to prediction of heart
Table 1
Correlation Coefficients of Various Parameters of
Cardiac Volumes in 261 Cases
Correlation coefficient (r) *
Variables
-0.032
-0.036
0.913
0.904
0.960
0.954
Race
Sex
Age
Height
Weight
Body surface area
*Product moment
or point
biserial
as
appropriate.
NGHIEM ET AL.
512
volume when weight was also used as an independent variable. By contrast, consideration of the sex of the subjects improved the
prediction of the cardiac volume by weight
(table 2). Males and females had different cardiac volumes predicted from weight
(t 6.03, P < 0.005) making the adoption of
two regression lines necessary.
Best fit regression equations relating cardiac volume to body surface area on the one
hand and to body weight on the other are
shown in table 3. The 95% confidence limits
Table 2
Normal Cardiac Volume for Male and Female Subjects as Predicted by Body Weight
Body weight (X)
in pounds
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
5
10
15
20
25
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
Mean
Cardiac volume (Y) of males (ml)
95% confidence limits
Upper
Lower
limit
limit
37
72
27
53
79
104
129
154
204
20
40
59
77
95
114
151
187
224
260
296
332
368
404
439
Cardiac volume (Y) of females (ml)
95% confidence limits
Upper
Lower
Mean
limit
limit
29
54
39
73
105
21
39
57
74
91
107
77
141
101
137
175
123
167
209
145
197
277
189
256
254
344
231
314
303
411
273
370
352
477
314
426
401
543
354
481
450
610
394
535
498
676
434
589
547
741
473
642
595
807
512
695
644
873
475
550
748
692
938
510
588
800
740
1004
626
545
852
788
1069
580
664
903
836
1134
616
702
954
For males Y 5.620 XO973 (linear corrEelation r -0.982)
For females Y - 6.628 X° 907 (linear correlation r - 0.980)
107
139
170
201
231
261
290
319
348
377
405
433
461
489
516
Table 3
Comparison of Body Surface Area and Body Weight as Predictors of Cardiac Volume
Sample
No.
of
subjects
Correlation
coefficient
(r)
95% confidence limits
Lower
% of mean
limit
Regression
equation
Upper
limit
Y=284 Z
= 295 Z
=273 Z
413 Z
434 Z
388 Z
Body surface area (Z)
Limited data
Males
Females
261
132
129
Limited data
Males
Females
Total data
Males
Females
261
132
129
305
158
147
0.981
0.980
0.987
Body weight (X)
0.987
0.990
0.990
0.985
0.985
0.989
Y=
=
=
Y
Y
Y
5.0X
5.2X
4.8X
4.9X
5.1X
4.7X
6.6X
6.7 X
6.4X
6.6X
6.8 X
6.3X
Z
Z
Z
+45
±47
±42
3.4X
3.7X
3.2X
3.2X
3.5X
3.1X
±32
±29
+34
+34
±32
±34
155
156
159
The limited data included 261 subjects whose height and weight were known. Y = cardiac volume in milliliters;
X body weight in pounds; Z = body surface area in square meters as computed from the DuBois and Dubois
formula.
Circuiation, Volume XXXV, March 1967
CARDIAC VOLUME
513
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
are 32 and 34% above or below the cardiac volume predicted from body weight in males
and females, respectively (figs. 2 and 3), as
compared with 47 and 42% above or below
the predicted value when body surface area
is employed. The merits of various formulae
correlating cardiac volume with body weight
are depicted by correlation coefficients in
table 4 (polynomial equations to the second
and third power with lower correlation coefficients are not listed). Linear regression
equations without intercept have the largest
correlation coefficients. Separating males and
females, the correlation coefficient improves
slightly for females and deviations of the confidence limits are closer to the mean for
males.
Evidence of reproducibility of the simplified method of cardiac volume determination
is shown in table 5. Using the measurements
NORMAL FEMALES
+1.97-
900
-
n = 147
Y 4.70 X
r = .989
Mean
700-
_s 600
00
-1.97.
:30
200
O
0
20
30
40
S0
60
70
80
90
100
110
120
130
140
150
160
170
Weight in lbs.
Figure 3
NORMAL MALES
Scattergram of cardiac volume versus body weight in
normal female subjects. The regression line (predicted
mean) and the 95% confidence limits (+1.97 sD or
+34% of the mean) are shown.
of observer I (M.H.S.) as a basis for comparison, the mean difference for a single
cardiac volume determination by each of the
other two observers was 3% (range 0 to 14%),
and the maximal difference less than 8% in
95% of the cases. Corelation coefficients in
regression of cardiac volume on weight were
high in spite of rater bias and random error
of measurements.
E
c0
o
Cardiac Volume (CV) and Cardiothoracic
Ratio (CTR) in Rheumatic Mitral Insufficiency
in Children
10
20
30
40
50
60
70
80
90
100
110
120
130
140
500
160
170
Weight in lbs.
Figure 2
Scattergram of cardiac volume versus body weight in
normal male subjects. The regression line (predicted
mean) and the 95% confidence limits (+1.97 SD or
+32% of the mean) are shown.
Circulation, Volume XXXV, Marcb 1967
CTR, as defined, is a relative measurement
comparing the horizontal projection of the
heart in the frontal plane, with the largest
diameter of the chest, used as a correction
factor for body size and magnification. Thb
merit of CTR as compared with cardiac volume index is shown here to be independent
of growth and stature.
The best fit regression equation of CVI on
CTR was a power function. Sex difference in
the prediction of CVI from CTR (P < 0.001)
NGHIEM ET AL.
514
as a warning sign of an imminently fatal
course. Considering for example a CTR in the
range of 0.65 ±40.02, two patients died, five
others had a progressively downhill course
from this point, and three recovered.
In 27 patients with pure rheumatic mitral
insufficiency grouped according to severity,'8
trends of cardiac volume change are depicted
in figures 5 through 8.
necessitated the adoption of two regression
equations (table 6). The corresponding regression lines are shown in figure 4.
In plotting CTR against time for the purpose of correlation with the clinical course,
trends of change in heart size were noted
in spite of relatively large fluctuations, similar to case 9 in an earlier report.25 No limits
of CTR, however, could be safely proposed
Table 4
Formulae for Prediction of Cardiac Volume by Body Weight
Correlation
coefficient
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Subjects
No.
Regression equations
Males
158
Y-5.1X
Y =4.6X+26
0.985
0.956
-
Y = 5.62 X 0-973
0.983
Y - 89 Exp.015OX
Y 4.7X
Y = 3.9 X + 33
0.890
0.989
0.965
Y
6.67 X 0.906
0.981
Y
Y
78 Exp.0163X
4.9X
0.895
0.985
Upper
limit
(r)
95% confidence limits
Lower
% of mean
limit
3.5 X
6.8X
± 32
+36
-26
Females
147
6.3X
3.1 X
+
34
+36
-26
305
Total
6.6X
3.2X
+
34
Abbreviations: Y = cardiac volume in milliliters; X = body weight in pounds.
Table 5
Prediction of Cardiac Volume by Body Weight
ments by Three Observers
Correlation
coefficient
No. of
cases in
as Derived
Regression
equations
from Independent Cardiac Volume Measure95% confidence limits
Upper
Lower
limit
limit
Deviations of
confidence limits
(as % of mean)
Observer
samples
(r)
I (M.H.S.)
305
0.985
Y
X
6.62
X
3.24
X
+34
261*
0.987
6.60
X
93
0.988
Y=4.99 X
Y 4.84 X
6.59 X
3.38 X
3.10 X
+32
±36
93
93
0.988
0.988
Y=4.86 X
Y
4.91 X
6.54 X
6.61 X
3.18 X
3.20 X
+35
+35
II
III
Abbreviations: Y
*The number of
according to sex.
4.93
cardiac volume in milliliters; X = body weight in pounds.
cases in which
both body weight and surface area were known. The samples
were
not grouped
Table 6
Relation of Cardiothoracic Ratio to Cardiac Volume Index in Children and Adolescents
with Rheumatic Mitral Insufficiency (Twelve Males and Fifteen Females)
Sex
No. of
determinations
(n)
Regression equation
Males
Females
71
108
CVI - 0.0032 CTR3.053
CVI - 0.0564 CTR2.278
Correlation
coefficient
(r)
0.920
0.939
Explained
variance
(r2)
p
0.85
0.88
< 0.001
< 0.001
Circulation, Volume
XXXV, March 1967
515
CARDlAC VOLUME
cardiac volume by weight. In the presence
of acute carditis, heart size exceeded twice
the upper limit of normal volume in two patients. Four patients had had congestive failure but after its control the clinical course
did not deteriorate. In two subjects, repeated
attacks of rheumatic carditis with congestive
heart failure responded well to salicylate and
steroid therapy.
C-diac Voluj
Index in ml/
3OA
IB80
1200
1000
Group 3 (Digitalized Cases of Severe Mitral
Insufficiency and Symptoms with Mild Exercise
at Rest)
-
Boo
600
Moles
*
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Io
.
.
CVI
= 71
.0032 CTR 3 -053
=
,2=
400
.920
.85
-
o
F-.l.s
CVI
=
.0564 CTR
r
=
.939
n ' 108
,2
20D
2.278
.88
This group included four males and six
females (figs. 7 and 8). Heart volume exceeding two times the upper limit of confidence was noted in all except one. Progressive cardiac enlargement occurred in this
patient although there was no evidence of a
GROUP
Il
40
50
60
so
70
CARDIOTHORACIC RATIO IN
90
PERCENTAGE
2000
Figure 4
Regression lines relating the cardiac volume index
(CVI) to cardiothoracic ratio (CTR) for male and
female patients with rheumatic mitral insufficiency.
n = number of determinations of CVI and CTR; r =
linear correlation between CVI and power transformation of CTR; and r2 = explained variance.
Group
1
or
(Asymptomatic with Mild Mitral
Male
90
2(m+1.97.)
a Inactive Ri. Fever
o C Initial attack of Rh. Fever
*
Repeat attock of Rh. Fever
o
Female
2(m+1.97.)
1500
E
Insufficiency)
This group included six males and six females (fig. 5). Characteristic of this group
was the normal cardiac volume except for
one patient whose heart volume was 6% above
the largest value predicted by weight and who
had had several attacks of carditis. Cardiac
volume during the acute phase of rheumatic
fever was normal or nearly normal in six subjects and moderately increased in three others.
Only two patients had mild congestive heart
failure with the first attack of carditis but
the cardiac enlargement never exceeded twice
the upper limit of normal volume.
Group 2 (Moderate Mitral Insufficiency with Signs
of Intolerance to Moderate or Marked Exertion)
This group consisted of two males and three
females (fig. 6). In all but one patient the
heart was 9 to 70% above the largest predicted
Circalation, Volame XXXV, March 1967
Mate
0
E)
(+1.97.)
1000 '
1.
Female
(4197-0
Male
(-197.1
500
-
Femole
(-1.97.1
U e , *4
0
30
40
50
60
70
80
90
100
110
120
130
140
150 160
Weight in lbs.
Figure 5
Trends of cardiac volume in male and female children
with mild mitral insufficiency (group 1). During the
acute attack of rheumatic fever, heart volumes were
normal or moderately increased (less than twice the
upper limit of normal). Subsequently cardiac volume
returned to normal in all but one patient who had
several attacks of rheumatic carditis.
NGHIEM ET AL.
516
GROUP 2
2000
Male
2(m+1.97s)
O O
O
0
Inactive Rh. Fever
Initial attack of Rh. Fever
Repeat attack of Rh. Fever
Femole
2(m+1.97*)
1500
E
>
Male
1.97
1000
Fermole
(+1.970)
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
SOO
The low incidence of active carditis during
terminal episodes of heart failure (two
of seven cases) may well be explained by
the exclusion of cases in which rheumatic activity was thought to be a major factor in
death.18 Myocarditis was not excluded in one
patient who died suddenly at home after progressive cardiomegaly for 17 months and
atrial fibrillation for at least 22 months. In
the remaining six cases, the course ranged
from the rapid demise of a female with
cardiac volume of 1,580 ml (fig. 8) to intractable congestive failure in another during
the last 20 months of her life. Several steroid
courses were ineffective in this and two
other cases in the terminal months.
Male
(-1.97-)
500
Female
-
GROUP 3 - MALES
1.97')
[it
0
2000-
7J
30
50
70
ilO
90
Weight
in
130
150
lbs.
Figure 6
Trends of cardiac volume in male and female children
with moderate rheumatic mitral insufficiency (group
2). During the acute attack of rheumatic fever, enlargement of the heart up to twice the upper limit
of normal volume was observed. In years of follow-up,
increase in cardiac volume (9 to 70% over the upper
limit of normal) was noted in all except one patient.
1500
-
E
,,
1000
.2
2
second rheumatic attack. Similar but more
rapid clinical and hemodynamic deterioration
occurred in the two surgically treated patients.
Using the usual clinical and laboratory criteria, the downhill course in these two cases
was not attributed to active carditis. At surgery no pericardial effusion was noted. Left
atrial biopsies showed "late phases"22 in one
and were not diagnostic in the other. Unexpectedly low stroke indices (25.4 and 33.4
ml/beat/m2) as compared with normal values,26 1 month postoperatively coincided with
cardiac volumes of 86 and 96% above upper
limits of normal. Normal stroke indices were
observed later when cardiac volumes were
almost normal.'s
500
1
t Deoth
o t /
30
50
70
90
It0
130
150
Weight in lbs
Figure 7
Trends of cardiac volume in male children with severe
rheumatic mitral insufficiency (group 3). Progressive
cardiac enlargement contrasts with reduced weight
gain. The recovery of the patient whose mitral valve
was replaced is striking both in terms of rapid weight
gain and diminution in cardiac volume. (Open square
indicates inactive rheumatic fever.)
Circulation, Volume XXXV, March 1967
CARDIAC VOLUME
517
GROUP 3- FEMALES
Initial attack of Rh. Fever
Repeat attack of Rh Fever
o Terminal heart
failure
* Mitral valve replacement
O
2000 -
t
1500
Death
/ 2(m+1.97.)
-
E
E4
000l-
+ I.97.
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
500
-
-1.9?.
0
H
/
O
30
50
70
90
110
30
5I0
Weight in lbs.
Figure 8
Trends of cardiac volume in female children with
rheumatic mitral insufficiency (group 3). Cardiac enlargement contrasting with reduced weight
gain in this group is similar to male children of group
3 (fig. 7). During the recovery of one patient whose
mitral valve was replaced, reduction in cardiac volume
occurred simultaneously with weight gain. (Open
circle indicates inactive rheumatic fever.)
severe
With due consideration of possible pericardial effusion, when cardiac volumes reached
1,100 to 1,300 ml, one patient died suddenly
and nine others followed a progressively downhill course. Rapid and severe loss of weight
(even in spite of edema), together with a steep
rise in cardiac volume, was one of the earliest
indicators of the terminal decline. Limits of
radiologically determined cardiac volume near
the time of death were 1,600 to 2,200 ml in six
cases. The weights of two hearts were 395
and 520 g, respectively, but radiologically
estimated cardiac volumes were 1,580 and
2,200 ml, the second value being augmented
by a 200 ml pericardial effusion. Pathological
diagnoses were compatible with "middle
Circulatino, Volume XXXV,
March 1967
phases" in the former and "late phases" in
the latter.22
Discussion
Limitations of the Radiological Ellipsoidal
Technique of Determining Cardiac Volume
Minor variations in radiological technique
produce insignificant changes in results,26 especially if the target-to-film distance is long
enough, is held constant, and the patient is
close to the film.
One possible source of error resides in extending the ellipsoidal approximation to irregular heart contours. In the case of enlarged
hearts tending to have two or three equal
diameters, the present assumption remains
valid. Compared with other techniques, the
ellipsoidal method results in a slight underestimation27 but results were within 5% of the
displaced cardiac volume for cadavers'
hearts.28 A constant magnification factor
(M) of
for films exposed at a distance
of six feet (183 cm) from the target, was
found suitable for a large number of persons.
For small infants with an anteroposterior distance of 7 cm and lateral distance of 7 cm
from the center of the heart to the film, the
basic constant may result in up to 6% underestimation of cardiac volume. Comparing all
our data with those of Axen and Lind29 who
measured the heart volume in 45 infants by
an elaborate technique, we found that the
95% limit of the total group in the present
study (Y = 6.6 X) accounted for 41 cases
(or 91%). A single constant was therefore
used for convenience.
Variations in cardiac volume may occur
during change from the recumbent to the
erect position,30 from diastole to systole when
the heart rate is slower than 70 beats per
minute, or during phases of respiration.31' 32
In unfavorable circumstances,26 these factors have been found to amount to less than
12% error at a confidence limit of to.o. Nevertheless, consecutive exposures at right angles,
which rarely fall precisely in end systole or
end diastole, the relatively fast heart rate
in young subjects, and a standardized practice of exposing films in moderate inspiration
518
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
assure reasonable reproducibility of the radiological technique.
Although observer variation resulted in no
more than an 8% error in 95% of the cases,
a value of 10% is suggested as an index of
true change of volume for enlarged hearts
because of ease of estimation. Rater consistency falling within 4% of the mean26 is not
different from measurements by separate
readers.
A drawback of the method is that estimation of cardiac volume includes the pericardial space and the contents of adjoining
vessels.33 For a given patient, rapid change
of adjacent vascular volume is not expected
to be large over short periods of observation.
Rheumatic pericardial effusion is rarely
enormous,34 nor does it persist in spite of
antirheumatic treatment34 or cause symptoms35
for months or years as in cases of severe
cardiomegaly. When in doubt, serial determinations of cardiac volume, angiocardiography,
radioactive scanning, or a diagnostic tap will
help in detecting an effusion.
Characteristics of the Study Sample
Selection of normal chest roentgenograms
as controls may exclude extremes of normal
heart size. However it includes the normal
limits as commonly agreed upon by radiologists and cardiologists.
Originally subjects were chosen to cover
adequately all age groups (fig. 1). For the
first 2 years of life, a larger number were
selected to cover the period of rapid growth.
A better correlation of cardiac volume with
body weight than with age cannot be attributed to choice of sample. Electrocardiograms
in 36 subjects (12% of total) were normal.
Obese,36 anemic37 and pregnant38 patients
were excluded because of likelihood of hemodynamic or blood volume disturbances.
Choice of an Independent Variable in
Predicting Cardiac Volume
The use of age as a predictor of cardiac
volume in children is based on the supposition
that growth is normal. This postulate does
not hold in the presence of heart disease.
A heart size "normal for age" may be exces-
NGHIEM ET AL.
sively large if the body fails to grow. Other
objections include poor correlation of age with
heart volume over a longer life span,'2 and
aging4 which introduces factors unrelated to
the postulate.
Sex difference in cardiac volume has been
a uniform finding in all series comparing males
and females", 12,19 except in Lind's study14 in
which an overlap of cardiac volume in both
sexes was to be expected in view of the small
body weight (292 infants below 12 kg).
Normal heart volume in girls as compared
vith boys appeared neither smaller from 2 to
7 years nor larger after 12 to 13 years. This
was in contrast with an earlier report based
upon differences in cardiothoracic ratio.39
The choice of either body weight or body
surface area as a predictor of cardiac volume
is most difficult. In reported series4 6 '0, 1 14, 40
neither has superior correlation with heart volume. An apparent paradox is their low correlation in some series.", 12 The explanation can
usually be found in reviewing these samples.
The magnitude of a correlation or regression
coefficient depends upon the range of values
for the independent variable. Selection of any
limited band of body weight or surface area
will cause correlation to drop sharply because
of restriction in range for the variable. Extrapolation of results from such samples becomes
questionable.
A universal finding however is that height
does not correlate well with cardiac volume.
Its use as an adjunct to weight does not
improve the prediction of the latter.4' In the
present study, predicted normal limits of cardiac volume by body weight are closer to the
regression lines (29 to 34%) than limits predicted by surface area. The use of weight is
more precise and eliminates an additional
measurement (height). In heart disease, the
ratio of cardiac volume to weight might provide a more sensitive index than the CVI.42
The advantage of linear regression equations
without intercept is evident( YX - K) . The 95%
confidence limits were used instead of standard deviation (a-) because the latter should
Circulation, Volume XXXV, March 1967
CARDIAC VOLUME
vary in absolute value with weight itself
(o'=-a' X). A clear concept of cardiomegaly
is obtained by comparing the excess in volume
with the upper limit of normal.
Cardiothoracic Ratio and Cardiac Volume
in Mitral Insufficiency in Children
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Studies of CTR in normal children and
adolescents43' 44 showed that normal values
were less than 0.50, and the range of normal
variation within 0.10 in the same subjects44
(for example, range 0.40 to 0.49). In the case
of cardiac enlargement secondary to rheumatic mitral insufficiency, CTR was less than
twice the upper limit of normal in male or
female subjects, whereas CVI was greater than
four times the normal upper limit in both
sexes (fig. 4). CTR must necessarily be related to CVI in a logarithmic progression.
From CTR values of 0.40 to 0.80, the slope of
the regression line becomes two times steeper
for females. In males, the slope ascends so
rapidly that maximum CVI is reached before
CTR is in the range of 0.80. The random
CTR error of ±0.05 as in normal subjects
corresponds to large variations in cardiac
volume when CTR is above 0.60 or 0.70. CTR
is thus a poor indicator of cardiac enlargement where accuracy is most needed (fig.
9). Unexplained variance of 12 to 15% is shown
by values of r2 in table 6. Weight loss, for
example, is accompanied by increased relative
cardiac volume but not by a change in the
maximal thoracic diameter.
CTR is generally less than 0.50 in normal
subjects and variations between 0.40 and 0.50
enable no conclusions to be drawn about differences in cardiac volume between males and
females. It is suggested, however, that, beyond
a CTR of 0.60, the more reliable technique
of cardiac volume determination is preferable
(fig. 4).
Significance of Cardiac Volume in
Rheumatic Mitral Insufficiency
Progressive cardiac enlargement culminating in death in the first two decades is not
unique in the present study. Similar observations were made by Noonan and Spencer,45
but volumetric limits were not stated.
Circulation, Volume XXXV, March 1967
519
The close correlation between cardiac enlargement and the clinical course of rheumatic
mitral insufficiency requires consideration of
the underlying mechanism. True cardiac enlargement was confirmed in four group 3 patients in whom the pericardia were opened.
The total cardiac weight in two autopsied
cases was no more than 25% of the radiographically determined cardiac volume. A
significant fact is that with severe cardiac
enlargement there are no detectable volume
changes during the cardiac cycle.46 Presumably, the right and left forward stroke volumes
diminish relative to increased cardiac volume and fall within the 5% error of the ellipsoidal approximation technique. For this reason, an increase of ventricular residual volume
compared with diminution of forward stroke
volume is not considered to be responsible
for the progressive cardiac enlargement detectable by the technique.
The mitral regurgitant volume during systole subsequently returns from the left atrium
to the left ventricle. In either diastole or systole, the left heart volume is augmented by
this amount which is accompanied by a proportional increase in the left ventricular
residual volume.47 Atrial residual volume may
be large in atrial fibrillation. Progressive enlargement of the left heart occurs if the
regurgitant stroke or residual volumes are
augmented. Fibrosis and retraction of the
mitral valve or chordae tendineae may increase the regurgitation. Further increase of
residual volume in the left ventricle or atrium
may lead to dilatation of these chambers
and perpetuate a vicious cycle.48' 49 Tricuspid
insufficiency may result with right ventricular
hypertension secondary to left heart failure or
increased pulmonary vascular resistance.50
Rapid increase in heart volume during the
terminal decline may be accentuated by tricuspid regurgitation which was recognized in
some cases in this study.
Comparing the three groups of patients
(figs. 5 to 8), the results of the present
study support the concept that the future
course of rheumatic mitral insufficiency is
determined by the severity of the initial
520
5NGHIJEM ET AL.
attack.5' In the presen-ce of cardiac enlargeinent during the acute phase, recovery wvas
characterized by diminution of cardiac volume.
In grouLp 2 patients, significant mitral re-
gurgitationxwas reflected by increased cardiac
volume xwbhich remained relatively proportional to growth for many years. The mechanical disadvantage in grouip 3 patients was such
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Figure 9
Cardiac volutmie determinationis int tnto patients withoutt active cat!ditis. The upper chiest
roentgenograims are front a patient in gronyi) 2, 2 years after the initial attack of rheuimatic
carditis. The hieart, was niodeatehly inicreased int volumre anid the clittical coturse was stable.
Despite a severe attack of rheumatiic carditis with con gestivc failure in September 1955, this
patienit was able to 1lay two sets of teinnis 11 yeairs later. The lower chest roentgenogyrams are
from a patient in group 3, 16 months aIfter the iniitial attaclk of rheumantic carditis. Pr-ogressive
enlargement of the hcart wcas associated wcithi a tapid cliniical dlclinte wvithioult detectable acutte
rheumatic carditis. Replacement of the mitral valve resulted int a rapid reduction of cardiac
voltume. (See fig. 7.) Similar cardiothloracic ratios in these two patients cdid not indlicate the
magnitude of cardiac enlargemenit or permit prediction of different clinical courses.
Circulation, Volume XXXV, March 1967
CARDIAC VOLUME
that recovery did not occur even in the
absence of demonstrable active carditis. Mitral
valve replacement in two patients was followed by a gradual return of resting cardiac
output to normal levels.8 Mitral insufficiency
per se appears to be at least the major factor
in hemodynamic deterioration and its early
correction is desirable for individuals in
group 3.
Acknowledgment
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
The authors wish to acknowledge the assistance
of J. E. Overall, Ph.D. and R. G. Kleinbrink, M.S.,
from The Research Computation Section, for statistical analyses, Alvin E. Rodin, M. D., Department of
Pathology, and H. H. Hammen, B.S., Pediatric
Cardiology Laboratory.
References
1. ROHRER, F.: Volumbestimmung von Korperhohlen und Organen auf orthodiagraphischem
Wege. Fortschr Rontgenstr 24: 285, 1916.
2. KAHLSTORF, A.: Vber eine orthodiagraphische
Herzvolumenbestimmung. Fortschr Rontgenstr
45: 123, 1932.
3. REEVES, T. J.: Year Book of Cardiovascular and
Renal Disease. Chicago, Year Book Medical
Publishers, 1963-1964, p. 498.
4. STRANDELL, T.: Heart volume and its relation
to anthropometric data in old men compared
with young men. Acta Med Scand 176: 205,
1964.
5. KEATS, T. E., AND ENGE, I. P.: Cardiac mensuration by the cardiac volume method. Radiology 85: 850, 1965.
6 LYSHOLM, E., NYLIN, G., AND QUARNA, K.:
Relation between heart volume and stroke
volume under physiological and pathological
conditions. Acta Radiol (Stockholm) 15: 237,
1934.
7. LILJEST RAND, G., LYSHOLM, E., NYLIN, G., AND
ZACHRISSON, C. G.: Normal heart volume in
man. Amer Heart J 17: 406, 1939.
8. COMEAU, W. J., AND WHITE, P. D.: Evaluation of heart volume determinations by the
Rohrer-Kahlstorf formula as a clinical method
of measuring heart size. Amer Heart J 17:
158, 1939.
9. COMEAU, W. J., AND WHITE, P. D.: Critical
analysis of standard methods of estimating
heart size from roentgen measurements. Amer
J Roentgen 47: 665, 1942.
10. BI6RCK, G.: On the relationship between the
heart volume and various physical factors.
Acta Radiol (Stockholm) 25: 372, 1944.
Circulation, Volume XXXV, March 1967
521
11. KJELLBERG, S. R., RUDHE, U., AND SJ6STRAND,
T.: Relation of the cardiac volume to the
blood volume and the physical capacity for
work. Acta Radiol (Stockholm) 31: 113,
1949.
12. MAUREA, G. NYLIN, AND SOLLBERGER, A.: Normal Heart volume. Acta Cardiol (Brux) 10:
336, 1955.
13. NYLIN, G.: Roentgenographic heart volume determination. In Examination of the Cardiac
Patient, edited by A. A. Luisada. New York,
McGraw-Hill Book Co., 1965, p. 4-365.
14. LIND, J.: Heart volume in normal infants. Acta
Radiol (Stockholm) (suppl. 82): 1950.
15. FRIEDBERG, C. K.: Diseases of the Heart, ed. 2.
Philadelphia, W. B. Saunders Co., 1956, p. 101.
16. STUART, H. C.: Percentile charts for height and
weight of normal infants and children (from
birth to 2 years and a half and from 2 years
to 13 years). Evansville, Indiana, Mead
Johnson & Co.
17. FALKNER, F.: Some physical growth standards
for white North American children. Pediatrics
29: 467, 1962.
18. HARRIS, L. C., NGHIEM, Q. X., AND SCHREIBER,
M. H.: Rheumatic mitral insufficiency in children: Course, prognosis and effect of mitral
valve replacement. Amer J Cardiol 17: 194,
1966.
19. SCHREIBER, M. H.: Volume of the heart: Roentgenographic determination. Arizona Med 21:
551, 1964.
20. DuBois, D., AND DuBois, E. F.: Formula to
estimate the approximate surface area if height
and weight be known. Arch Intern Med
(Chicago) 17: 863, 1916.
21. Jones Criteria (Revised). New York, American
Heart Association, 1965.
22. GROSs, L., AND EHRLICH, J. C.: Studies on the
myocardial Aschoff body: II. Life cycle, sites
of predilection and relation to clinical course
of rheumatic fever. Amer J Path 10: 489,
1934.
23. NADAS, A. S.: Pediatric Cardiology, ed. 2.
Philadelphia, W. B. Saunders Co., 1963, p. 39.
24. SNEDECOR, G. W.: Statistical methods, ed. 5.
Ames, Iowa, Iowa University Press, 1964,
p. 413.
25. TAUSSIG, H. B., AND GOLDENBERG, M.: Roentgenologic studies of size of heart in childhood:
I. Three different types of teleroentgenographic
changes which occur in acute rheumatic fever.
Amer Heart J 21: 440, 1941.
26. KJELLBERG, S. R., LONRoTH, H., AND RUDHE,
U.: Effect of various factors on the roentgenological determination of the cardiac volume. Acta Radiol (Stockholm) 35: 413,
1951.
NGJIIEM ET AL.
522
27. GEBHARDT, W., DANNER, D., REINDELL, H.,
AND
K6NIC, R.: Eine vergleichende Untersuichuing
zur methodischen Fehlerbreite der rontgenologisehen Herzvolumenbestimmung. Acta Med
Scand 167: 467, 1960.
28. FRIEDMAN, C. E.: Residual blood of the heart.
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Amer Heart J 39: 397, 1950.
29. AxELN, O., AND LIND, J.: Roentgenologic determination of heart volume in infants. Acta
Paediat (Stockholmi) 32: 270, 1945.
30. LAsISSON, H., AND KJELLIIERC, S. B.: Roentgenological heart volume determination with
special regard to pulse rate and the position
of the body. Acta Radiol (Stockholm) 29:
159, 1948.
31. SPROUL, A., AND SIMPSON, E.: Stroke volume
aind related hemodynamnic data in normal children. Pediatrics 33: 912, 1964.
32. NATVIlC, P.: Vonltiiie of the heart in Miiller's
and Valsalva's tests. Acta Radiol (Stockholmn)
15: 657, 1934.
33. TIURN, P.: Zur r6intgenologischen Voluimenimiessutng des HIerzens. Fortschr lontgenstr 90:
290, 1959.
34. MARKOWITZ, M., AND KUTTNER, A. C.: Rheumatic Fever: Diagnosis, Management and Prevention. Philadelphia, W. B. Saulnders Co.,
1965, p. 45.
35. WINTERIS, W. L., AND SOLoFF, L. A.: Pericardial
effusion in clinically inactive compensated
rheumatic heart disease. Amer J Mecl Sci
242: 173, 1961.
36. CAYLE11, C. C., MAYS, J., AND RILEY, HI. D., Jni.:
Cardiorespiratory syndromie of obesity (Pickwickian syndromie) in children. Pediatrics 27:
237, 1961.
37. BRIANNON, E. S., MERRILL, A. J., WARREN, J. V.,
AND STEAD, E. A., Jn.: Cardiac output in
patients withl chlronic anemia as measured by
the techniiquie of right atrial catheterization.
J Clin Inivest 24: 332, 1945.
38. KJELLRERC, S. R., L6jNISOTH, H., RUD}HE, U.,
ANfD SJ6STRAND, T.: Blood voluIme and heart
volume during pregnancy and the puierperitm.
Acta Med Scand 138: 421, 1950.
39. LINCOLN, E. M., AND SPILLMAN, R.: Studies
on the hearts of normal children. Amer J Dis
Child 35: 791, 1928.
40. CARLCRIEN, L. E.: Gallop rhythm in children.
Acta Paediat (Stockholm) 33 (suppl. 6):
1946.
41. LuDwIc, H.: Rdntgenologische Beurteilung der
Herzgriisse. Fortschr Rontgenstr 59: 1, 1939.
42. DOMENET, J. G., EVANS, D. W., AND HOWARTJI7
F. H.: Clinical experience with radiological
determination of the heart volume. Brit Heart
J 25: 575, 1963.
43. ZISKINx, T.: Development and size of the heart
in children. Amer J Dis Child 30: 851,
1925.
44. MARESH, M. NI.: Growth of the heart relatedl
to bodily growth during childhood and adolescence. Pediatrics 2: 382, 1948.
45. NOONAN, J. A., AND SPENCER, F. C.: Surgical
treatmlent of rheumatic heart disease in childhood. Circuilation 30: (ssuppl. 3): 111-133, 1964.
46. EVANS, D. W., AND CARPENTER, P. B.: Errors
involved in radiological heart volume deter-
mination by the ellipsoid-approximation technique. Brit Heart J 27: 429, 1965.
47. JONES, J. W., RACKLEY, C. E., BRUCE, R. A.,
DoDc.E, H. T., COBB, L. A., AND SANDLER, H.:
Left ventricular volumiies in valvular heart
disease. Cireulation 29: 887, 1964.
48. LEVY, M. J., VAnRco, R. L., LILLEHEI, C. W.,
AND EDWARDS, J. E.: Mitral insufficiency in
infants, childlren, and adolescents: A review
of etiologic, electrocardiographic, radiologic
and pathologic factors, and surgical techniques.
J Thorac Cardiov Surg 45: 434, 1963.
49. LEvY, M. J., AND EDWARDS, J. E.: Anatomy
of mitral insufficiency. Progr Cardiov Dis 5:
119, 1962.
50. Ross, J., BRAUNWALD, E., AND MoRRow, A. G.:
Clinical and hemodynamiic observations in pure
mitral insufficiency. Amier J Cardiol 2: 11,
1958.
51. WOOD, P.: Diseases of the Ileart and Circulation, ed. 2. Philadelphia, J. B. Lippinicott Co.,
1956, p. 505.
Circulation, Volume XXXV, March 1967
Cardiac Volume in Normal Children and Adolescents: Its Application to Patients
with Rheumatic Mitral Insufficiency
QUANG X. NGHIEM, MELVYN H. SCHREIBER and LEONARD C. HARRIS
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Circulation. 1967;35:509-522
doi: 10.1161/01.CIR.35.3.509
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1967 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/35/3/509
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/