Download Print - Circulation

2028
Growth of the Great Vessels in the Normal
Human Fetus and in the Fetus
With Cardiac Defects
Philip C. Ursell, MD; Julianne M. Byrne, PhD; Thomas R. Fears, PhD;
Barbara A. Strobino, PhD; and Welton M. Gersony, MD
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Background. There is a paucity of quantitative anatomic data regarding human great vessel
development that could be useful as a reference for fetal echocardiographers who must
distinguish abnormal from normal cardiac development at early stages.
Methods and Results. To determine normal growth patterns, we plotted the diameters of the
aortic and pulmonary valves, ductus arteriosus, aortic isthmus, and descending aorta in 274
autopsy specimens from nonselected spontaneous abortuses of normal karyotype. There was a
linear increase in the diameters of these structures within the developmental period studied
(10-26 weeks). A relative narrowing of the aorta at the isthmus compared with the aortic valve
and descending aorta probably indicates that the majority of fetal left heart output goes to the
developing heart and brain. In contrast to previous studies of late gestation and neonatal
animals, however, we found that the diameter of the aortic isthmus was larger than that of the
ductus arteriosus, suggesting substantial isthmic blood flow in these midtrimester fetuses.
Among nineteen other hearts with diverse defects, both of two hearts with a narrow isthmus had
an enlarged ductus arteriosus and one heart with pulmonary atresia/intact septum had a
narrow ductus and increased aortic valve diameter.
Conclusions. During midgestation, the normal heart may have substantial aortic isthmic
blood flow that diminishes due to rerouting in late gestation when increased requirements of the
fetal brain and other organs prevail. Although fetal shunts may explain some vessel abnormalities, the majority of cardiac defects in this study were not associated with abnormal growth
of the great vessels within this developmental age range. (Circulation 1991;84:2028-2033)
A
lthough the development of the aorta and
pulmonary artery has been well studied in
human embryos as well as in animal models,
the rate of growth of these vessels during human
gestation has not been determined. The flow-dependence rule' states that the size of a vessel is proportional to the amount of blood flow through its lumen.
Therefore, cardiovascular malformations that result
in obstruction to blood flow or abnormal flow patterns may be expected to have an impact on the
development of the great vessels during fetal life. In
From the Departments of Pathology (P.C.U.) and Pediatrics
(W.M.G.), The Gertrude H. Sergievsky Center, Columbia Univer-
sity and the New York State Psychiatric Institute (B.A.S.), New
York, and the Epidemiology and Biostatistics Program, National
Cancer Institute (J.M.B., T.R.F.), National Institutes of Health,
Bethesda, Md.
At the time of this study Dr. Ursell was an Investigator for the
American Heart Association, New York City Affiliate.
Address for correspondence: Philip C. Ursell, MD, Department
of Pathology, University of California at San Francisco, San
Francisco, CA 94143-0506.
Received May 8, 1990; revision accepted July 2, 1991.
the current era of high-resolution echocardiography,
many cardiac defects are diagnosable in the fetus and
therapeutic options may be considered at a relatively
early developmental stage. Thus, it is important to
define the normal growth patterns for all cardiovascular structures to recognize abnormal development
as early as possible.
We studied the growth and development of the
great vessels in the fetus by examining cardiovascular
structures in autopsied human abortuses. The size of
the great vessels in consecutive (nonselected) spontaneous abortuses of normal karyotype without cardiovascular malformations was measured, and
growth curves were generated from the collected
data. To study the impact of cardiovascular malformations on growth of the fetal great vessels, we also
examined a small set of hearts containing cardiac
defects and compared the morphological features
with those in normal hearts. Our objectives were to
generate growth curves for normal development of
the great vessels in the human fetus and to examine
the relation between heart defects and growth of the
great vessels in utero.
Ursell et al Growth of Human Fetal Great Vessels
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Methods
All of the normal hearts examined in this study
were obtained from a large epidemiological, cytogenetic, pathological study of consecutive spontaneous
abortions identified at three New York City Hospitals over a 7-year period.2 Of 4,000 specimens collected, hearts from 412 fetuses at 10-26 weeks of
developmental age were identifiable. All were large
enough to evaluate anatomically.2'3 Of these 412
hearts, 277 were from fetuses of normal karyotype;
all except three were normal anatomically, and the
274 hearts served as the basis of this study. The aortic
and pulmonary valves were measured in each heart.
In addition, 173 of these specimens had the great
vessels attached so that the orientation was known
and measurements of the ductus arteriosus, aortic
isthmus, and descending aorta could be made. Nineteen complete specimens with cardiac defects were
analyzed separately as described below. In each of
the total 293 cases, a complete autopsy of the fetus
and placenta was performed.4 Developmental age
was determined by crown-to-rump length.5 The
hearts were relatively nonmacerated and preserved
in 10% neutral buffered formalin (4% formaldehyde)
for subsequent study.
Dissection of the fixed hearts and great vessels was
performed according to standard autopsy techniques.4 Each heart was opened along the path of
blood flow. All observations and measurements were
made on the preserved specimen under 10-fold magnification by one person (P.C.U.) using a dissecting
microscope. Previous studies have shown that there is
no discernible difference in measurements of great
vessel diameters after fixation in 4% formaldehyde
compared with the fresh specimen.6 Measurements
of the internal circumferences of the aortic and
pulmonic valve annuli, aortic isthmus, ductus arteriosus at its midpoint, and descending aorta one-half
centimeter distal to the ductus were made with a
small, flexible ruler or with a piece of thread and the
ruler. For the purpose of graphic presentation, the
diameter was calculated from the circumference,
assuming that the structures were approximately
circular in situ. For each parameter, the regression of
that heart measurement on developmental age was
obtained. Confidence intervals (68% and 95%) were
calculated.
To investigate the effect of various congenital
heart defects on the size of the arterial valves and the
great vessels including the ductus arteriosus, we
compared 19 malformed hearts (12 from induced and
four from spontaneous abortuses of known abnormal
karyotype as well as three from spontaneous abortuses of normal karyotype) with corresponding normal hearts from spontaneous abortuses of normal
karyotype. This set of malformed hearts of known
developmental age had been collected during the
same 7-year period. These hearts had been preserved
in formalin and dissected exactly as those in the
larger set had been. The above-described measure-
E
2029
688%
,,
,
4.0
3.0-
10
C
~~~~~~~~~~~~~~~~MEAN
cW__
5 c
11
12
13
14
15
16 17 18 19 20 21 22
DEVELOMNAL AGE on weeks)
23
24
25
26
FIGURE 1. Plot showing pulmonary valve diameter as a
function of developmental age. Among normal hearts, there is
a linear increase in pulmonary valve diameter with advancing
developmental age. Of 19 fetuses with cardiovascular defects,
one fetus with atrioventricular septal defect at 16 weeks had
significantly smaller pulmonaty valve diameter and one fetus
at 23 weeks had pulmonaty atresia and intact ventricular
septum. Cl, confidence limit; A, heart with ventricular septal
defect; *, heart with atrioventricular septal defect; x, heart
with narrow aortic isthmus; o, heart with pulmonaty atresia
and intact ventricular septulm.
ments were made in each heart. The morphological
characteristics of these fetal hearts were described
according to the scheme of Tynan et al.7 In addition
to clear-cut abnormalities such as septal defects,
more subtle changes caused by relative stenoses or
dilatations were considered to be abnormal if measurements were outside the 95% confidence intervals
of those of the normal hearts in the corresponding
developmental age range. For example, aortic isthmus narrowing was defined as those cases in which
the isthmus was smaller than the lower 95% confidence interval. Similarly, the ductus arteriosus was
considered to be enlarged or dilated if it had a
diameter greater than the 95% confidence interval.
The values for each parameter in the abnormal
hearts were plotted on the graphs generated from the
data obtained from the normal hearts.
Karyotypes were determined by tissue culture
of fetal tissue and subsequent staining by the Gbanding method.8
Results
Normal Hearts
Among the 274 normal hearts from spontaneous
abortuses of normal karyotype, there was a linear
increase in the diameters of the aortic and pulmonary
valve annuli, ductus arteriosus, aortic isthmus, and
descending aorta with advancing developmental age
(Figures 1-5). At each developmental age, the diameter of the aortic isthmus was greater than that of the
ductus arteriosus (Figures 3 and 4). Diagrams showing the 50th centile measurement at the above locations for each of four representative developmental
ages are presented in Figure 6. Ratios of ductus
arteriosus diameter to pulmonary valve diameter and
Circulation Vol 84, No
2030
5
November 1991
2.5-
4.5
4.0
wC i
--
_ _
A
_-
-
A----
'E
68%
Gr-
MA
MEAN
n
68% C
2.0-
U)
A---
95% c
I
96%CI
1.0 I
._
X
0.5
10
12
11
13
14
15
16 17 18 19 20 21 22
DEVELOPMENTAL AGE Cn weeks)
23
24
10
26
25
11
12
13
14
16 17 18 19 20 21 22
DEVELOPMENTAL AGE in week)
15
23
24
25
28
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
FIGURE 2. Plot showing aortic valve diameter as a function
of developmental age. Normal hearts show linear increase in
aortic valve diameter with advancing developmental age. One
fetus with narrow aortic isthmus at 18 weeks had significantly
smaller aortic valve diameter. Fetus with pulmonary atresia
and intact ventricular septum (23 weeks) had enlarged aortic
valve diameter. Cl, confidence limit; A, heart with ventricular
septal defect; *, heart with atrioventricular septal defect; x,
heart with narrow aortic isthmus; o, heart with pulmonary
atresia and intact ventricular septum.
FIGURE 4. Plot showing diameter of ductus arteriosus as a
function of developmental age. Linear increase in diameter of
the ductus arteriosus occurs throughout the second trimester of
normal fetal life. Both specimens with narrow aortic isthmus
had significantly enlarged ductus arteriosus, whereas specimen
with pulmonary atresia and intact ventricular septum had a
very narrow ductus. Cl, confidence limit; A, heart with
ventricular septal defect; *, heart with atrioventricular septal
defect; x, heart with narrow aortic isthmus; o, heart with
pulmonary atresia and intact ventricular septum.
aortic isthmus diameter to both aortic valve and
descending aorta diameters (Table 1) remained constant to the end of the second trimester.
aortic isthmus, ductus arteriosus, and descending
aortic diameters of these specimens are plotted individually on the normal growth curves (Figures 1-5).
Three of 12 specimens with a ventricular septal
defect and both of the specimens with a narrow aortic
isthmus and no other defect showed a ductus that
was definitely enlarged (i.e., above the 95% confidence interval for developmental age). The pulmonary annulus was enlarged in only one heart with
ventricular septal defect and one with atrioventricular septal defect but was not enlarged in either
specimen with a narrow aortic isthmus. The aortic
isthmus was normal in the three hearts with ventricular septal defect and enlarged ductus. The aortic
valve diameter in one of the two hearts with a narrow
Hearts With Cardiovascular Defects
There were 12 hearts with perimembranous ventricular septal defects, four with atrioventricular septal defects, two with a narrow aortic isthmus and no
other defect, and one with pulmonary atresia and
intact septum. The size of the ventricular septal
defect was smaller than the aortic root in all but two
cases; in these two hearts, the greatest diameter of
the defect was similar to the aortic valve diameter.
The measurements of the aortic and pulmonic valves,
3.0A
3.5 -
i
,-
2.5
C
-1
,---
c
_-6B
oum|L
C
MEAN
-
r 2.0
A
1.5-
--A
.-MEAA <
E
NCI
_
'
on
--95%C I
S6%CI
-,
3.0~~~~~~~~~~~~~~~
_8B ClI
,,-- 9% C
--
2.0
-
1.0-
_-
2 10~
,,
1.0 -
-
-
0
x
=c
v.v
10
11
-
12
..
,
,
, ,
-,
,-,
......................
,-
13
14
15
16 17 18 19 20 21 22
DEVELOPMALAGE (in week)
23
24
25
26
FIGURE 3. Plot showing diameter of aortic isthmus as a
function of developmental age. With advancing development,
normal hearts show linear increase in diameter of the aortic
isthmus. There were two specimens with significantly smaller
isthmuses and no other defect. Cl, confidence limit; A, heart
with ventricular septal defect; *, heart with atrioventricular
septal defect; x, heart with narrow aortic isthmus; o, heart
with pulmonary atresia and intact ventricular septum.
10
11
12
13
14
15
16 17 18 19 20 21 22
DEVELOPMENTAL AGE n weeks)
23
24
25
28
FIGURE 5. Plot showing diameter of distal aorta as a
function of developmental age. Among normal fetuses, diameter of the distal aorta increases linearly with advancing
developmental age. Cl, confidence limit; a, heart with ventricular septal defect; *, heart with atrioventricular septal defect;
x, heart with narrow aortic isthmus; o, heart with pulmonary
atresia and intact ventricular septum.
Ursell et al Growth of Human Fetal Great Vessels
<12 weeks
.16 weeks
2031
> 23 weeks
20 we.eks
DEVELOPMENTAL AGE
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
FIGURE 6. Fiftieth centile vessel diameters (in millimeters) at four developmental stages are shown diagrammatically. At each age,
pulmonary valve diameter is greater than that of the aortic valve. There is relative narrowing of the aorta at the isthmus, and
diameter of the isthmus is larger than that of the ductus arteriosus.
aortic isthmus was small. One heart with pulmonary
atresia and intact ventricular septum had a significantly smaller ductus diameter; the same specimen
had a large aortic valve. One of the four hearts with
atrioventricular septal defect had a significantly small
pulmonary valve.
Discussion
Our data show that in hearts from normal secondtrimester fetuses, there is a linear increase in the
diameters of the arterial valves and great vessels from
10 to 26 weeks. At any single developmental age,
pulmonary valve diameter was slightly larger than
aortic valve diameter. By the flow-dependence rule,1
this is consistent with at least equivalent blood flow
across the pulmonary valve compared with the aortic
valve during this stage of fetal development. Moreover, the diameter of the aortic isthmus was smaller
than that of either the aortic valve or the descending
aorta. Rudolph and colleagues9-11 noted this relative
isthmic narrowing in normal late gestational animals
as well as in human neonates. These investigators
attributed this finding to large cranial blood flow
from the ascending portion of the aorta, resulting in
less flow through the isthmus into the descending
aorta. They reported that the diameter of the ductus
arteriosus was greater than that of the aortic isthmus.
The data from the present study, however, were
obtained from hearts at an earlier developmental
stage. During midtrimester human fetal development, we found that the ratio of isthmus to descend-
ing aortic diameter was 71-78%, similar to the 75%
figure cited for normal human fetuses6 and neonates.'1 However, in contrast to studies performed in
late gestation,9-11 the mean diameter of the aortic
isthmus was larger than that of the ductus arteriosus.
A previous investigation using direct measurement of
fixed specimens also documented a larger isthmus
compared with ductus arteriosus in midtrimester
human fetuses.6 This observation, which argues
against the concept of large fetal blood flow through
the ductus arteriosus and relatively little through the
isthmus during midgestation, may have several explanations. First, the effect of cranial blood flow on
vessel size may not be reflected until a later gestational age. Indeed, it is during the last part of
gestation that blood flow to the brain and other
organs increases dramatically.10 During this latter
period, the growth of the ascending aorta and ductus
arteriosus may outstrip that of the isthmus, thus
accentuating the narrowing. Second, it could be said
that the ductus arteriosus in specimens from our
study and the previous investigation6 may be artifactually constricted due to fixation. In this regard,
freeze-fixation of the vessel may be preferable, as has
been demonstrated in neonatal rats.'2 We presume,
however, that the lesser amount of smooth muscle
and its relative immaturity would make the ductus
arteriosus from the midtrimester human fetus less
susceptible to the astringent formaldehyde used in
our study. This absence of constriction artifact during
midtrimester development is suggested by the con-
TABLE 1. Ratios of Mean Vessel Diameters at Four Developmental Stages
Developmental age (in weeks)
Ductus diameter/pulmonary valve diameter
Isthmus diameter/aortic valve diameter
Isthmus diameter/descending aorta diameter
< 12
16
36%
55%
35%
61%
71%
73%
20
36%
65%
71%
>23
40%
58%
78%
2032
Circulation Vol 84, No 5 November 1991
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
stant ratio of ductus arteriosus diameter to pulmonary valve diameter shown in Table 1. Third, our
results are supported by echocardiographic data from
living fetuses at 23-27 weeks of gestational age that
suggest at least equivalent size of the ductus arteriosus and isthmus in vivo.6 Thus, the animal models
from which the concept of high ductal-low isthmic
blood flow in the fetus was formulated may not
translate accurately to the human situation.
A particular strength of our study is the large
number of normal specimens used to generate the
growth curves. For the purpose of presentation, we
chose to use confidence intervals to define normal
versus abnormal measurements. We recognize, however, that some points close to but outside of the 95%
confidence interval may be at the end of a spectrum
of normal. For example, two fetuses had a narrow
aortic isthmus and no other defect. One examined at
a developmental age of 18 weeks had an aortic
isthmus diameter barely smaller than the 95% confidence interval. In this case, the small aortic valve
diameter and enlarged ductus arteriosus suggest a
significant albeit mild abnormality compared with the
other specimen at 21 weeks, which had an extremely
narrow isthmus consistent with severe coarctation.
Neither specimen had the juxtaductal ridge in the
aorta similar to the obstructing shelflike projection
found in neonates with coarctation. We presume that
this "shelf" is an acquired lesion related to contraction of smooth muscle in the wall of the ductus
arteriosus after birth. Our definition of aortic isthmus
narrowing is similar to if not more stringent than that
of others who have defined abnormal as being an
isthmic diameter no more than 40% of that of the
ascending aorta.13,14
Cardiovascular hemodynamics during embryogenesis has long been considered important in determining aortic arch selection (preservation of some arch
segments and involution of others) as well as normal
or abnormal cardiac morphogenesis. For example,
Rychter15 has shown experimentally that mechanical
obstruction of certain of the aortic arches in avian
embryos causes reproducible malformations of the
great vessels. Similar principles may be applied to
explain known cardiovascular malformations in humans. Rudolph and colleagues9 postulated anatomic
effects of a variety of defects on cardiovascular
structure and analyzed clinical data including angiograms from human neonates and young infants to
confirm that in cardiac lesions with reduced pulmonary arterial blood flow (e.g., pulmonary or tricuspid
atresia), the diameter of the aortic isthmus is increased, and that in cardiac defects with left ventricular outflow tract obstruction, the diameter of the
aortic isthmus is abnormally decreased. An example
of this concept is the hypothesis that leftward malalignment of the interventricular septum leads to the
development of hypoplasia of the aortic arch or
coarctation.13'16 Because none of our specimens contained a malalignment type of ventricular septal
defect, we could not examine the effect of this defect
in the midtrimester fetus. Our study did include one
heart with pulmonary atresia and intact septum that
had a significantly enlarged aortic isthmus; it also
included two specimens with aortic isthmus narrowing, both of which had an enlarged ductus arteriosus
suggestive of larger-than-normal flow across this vessel. These observations from a limited number of
specimens support the concept of an important hemodynamic influence on the structure of the developing human cardiovascular system. To our knowledge, this study is the first to show that hemodynamic
effects may be apparent in the human fetus as early
as 18 weeks of developmental age. It is important to
emphasize, however, that in the majority of specimens with cardiac defects, we found normal great
vessel dimensions. Thus, although the sample of
abnormal hearts was small, it is clear that not all
intracardiac defects are associated with abnormalities in the growth of great vessels in the human fetus
within this developmental age range.
The flow-dependence rule1 may reflect only one
of a number of elements associated with abnormal
growth of the great vessels. Cardiac dysgenesis may
also be an important factor. It has been shown that
complete removal of the cardiac neural crest in
chick embryos results in persistent truncus arteriosus; partial ablation of these tissues leads to a
variety of defects characterized by a malposed aorta
(e.g., double-outlet right ventricle, tetralogy of Fallot).17 There is circumstantial evidence that similar
pathogenetic mechanisms apply to human cardiac
maldevelopment.17
In this study, we characterized the growth pattern
of the great vessels in fetuses from 10 to 26 weeks of
developmental age. This period of human development follows completion of cardiogenesis and precedes the exponential phase of growth that occurs
during the last trimester. Although fetal heart size
at 15 weeks is at the limit of resolution of current
echocardiographic techniques, in the future this
and other clinical tools will be developed to have
the capability of investigating smaller-size organs.
Until now, there has been a paucity of quantitative
anatomic data regarding the early developmental
stages addressed by our study. The growth curves
generated from our data should be useful as a
reference for fetal echocardiographers who must
distinguish abnormal from normal cardiac development at early fetal ages.
Acknowledgments
We gratefully acknowledge Dr. Dorothy Warburton of Columbia University for the karyotype analyses, Zena Toran for artwork, and Lourdes M. Waters
for secretarial assistance.
References
1. Krediet P, Klein HW: Synopsis of normal cardiac development, in Pexieder T (ed): Perspectives in Cardiovascular
Research. New York, Raven Press, Publishers, 1981, pp 7-16
Ursell et al Growth of Human Fetal Great Vessels
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
2. Byrne J, Warburton D, Kline J, Blanc W, Stein Z: Morphology
of early fetal deaths and their chromosomal characteristics.
Teratology 1985;32:297-315
3. Ursell PC, Byrne JM, Strobino BA: Significance of cardiac
defects in the developing fetus: A study of spontaneous
abortuses. Circulation 1985;72:1232-1236
4. Byrne J: Fetal pathology laboratory manual. Birth Defects
1983;19:10-19
5. Potter EL, Craig JM: Rate of antenatal growth, in: Pathology
of the Fetus and the Infant, ed 3. Chicago, Year Book Medical
Publishers, Inc, 1975, p 19
6. Angelini A, Allan LD, Anderson RH, Crawford DC, Chita SK,
Ho SY: Measurements of the dimensions of the aortic and
pulmonary pathways in the human fetus: A correlative echocardiographic and morphometric study. Br Heart J 1988;60:
221-226
7. Tynan MJ, Becker AE, Macartney FJ, Jimenez MQ, Shinebourne EA, Anderson RH: Nomenclature and classification of
congenital heart disease. Br Heart J 1979;41:544-553
8. Warburton D, Stein Z, Kline J, Susser M: Chromosomal
abnormalities in spontaneous abortion: Data from the New
York City study, in Porter IH, Hook EB (eds): Human
Embryonic and Fetal Death. New York, Academic Press, Inc,
1980, pp 261-288
9. Rudolph AM, Heymann MA, Spitznas U: Hemodynamic
considerations in the development of narrowing of the aorta.
Am J Cardiol 1972;30:514-525
2033
10. Rudolph AM: The fetal circulation, in: Congenital Diseases of
the Heart. Chicago, Year Book Medical Publishers, Inc, 1974,
pp 2-9
11. Rudolph AM: Aortic coarctation and narrowing, in: Congenital Diseases of the Heart. Chicago, Year Book Medical Publishers, Inc, 1974, pp 330-346
12. Hornblad PY, Larsson KS: Studies on closure of the ductus
arteriosus: I. Whole-body freezing as improvement of fixation
procedures. Cardiologia 1967;51:231-241
13. Moulaert AJ, Bruins CC, Oppenheimer-Dekker A: Anomalies
of the aortic arch and ventricular septal defects. Circulation
1976;53:1011-1015
14. Blackburn W, Wheeler VR, Cooley NR Jr, Manci EA: Studies
of normal and abnormal aortic arch development in the fetus.
Proc Greenwood Genet Center 1987;6:112
15. Rychter Z: Experimental morphology of the aortic arches and
the heart loop in chick embryos. Adv Morphogenesis 1962;2:
333-371
16. Becker AE, Anderson RH: Malformations of the aortic arch,
in: Pathology of Congenital Heart Disease. London, Butterworths & Co Ltd, 1981, pp 321-338
17. Kirby ML, Waldo KL: Role of neural crest in congenital heart
disease. Circulation 1990;82:332-340
KEY WORDS * great vessels * fetal cardiac development
Growth of the great vessels in the normal human fetus and in the fetus with cardiac
defects.
P C Ursell, J M Byrne, T R Fears, B A Strobino and W M Gersony
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Circulation. 1991;84:2028-2033
doi: 10.1161/01.CIR.84.5.2028
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1991 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/84/5/2028
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/