Left ventricular remodeling in the first year after

acute myocardial infarction and the predictive

value of N-terminal pro brain natriuretic peptide
Jens C. Nilsson, MD,a Bjoern A. Groenning, MD,a Gitte Nielsen, MD,b Thomas Fritz-Hansen, MD,a
Jürgen Trawinski, PhD,c Per R. Hildebrandt, DMSc,d Gorm B. Jensen, DMSc,b Henrik B.W. Larsson, DMSc,a and
Lars Sondergaard, MDa Hvidovre and Frederiksberg, Denmark, and Basel, Switzerland

Background Left ventricular (LV) remodeling after myocardial infarction (MI) has received much attention because of
its severe impact on morbidity and mortality rates. However, the incidence and extent of LV remodeling in a modern infarct
population who were offered antiremodeling treatment in compliance with daily clinical practice is unknown. The purpose of
this study was to clarify this issue and to evaluate the predictive value of N-terminal pro brain natriuretic peptide (NT-
proBNP).
Methods Forty-two patients with a first transmural MI were examined after 1 week, 1 month, 3 months, 6 months, and 1
year with blood samples and magnetic resonance imaging.
Results In 12 patients (29%), LV end-diastolic volume index (LVEDVI) and LV end-systolic volume index (LVESVI)
increased by 24% and 22% (P <.0001; P = .01). In 12 patients (29%), LVEDVI and LVESVI decreased by 19% and 23% (P
<.0001; P = .0005), whereas the remaining 18 patients (43%) had stable conditions regarding these LV measures. LV ejec-
tion fraction at baseline was significantly reduced in all patient categories but was unchanged over time. Elevated NT-
proBNP level at baseline was identified as an independent predictor of increase in LVEDVI during follow-up examination (P
= .007). A baseline level of NT-proBNP >115 pmol/L identified patients who later had LV dilatation develop with a sensitiv-
ity and specificity of 89% and 68% (area under curve = 0.77).
Conclusion In this 1-year follow-up study of patients with a first transmural MI, approximately 30% had significant
increments develop in LVEDVI and LVESVI, and LV ejection fraction remained unchanged. Patients in whom LV dilatation
developed could be identified early after the MI with elevated plasma levels of NT-proBNP. (Am Heart J 2002;143:696-
702.)

Remodeling of the left ventricle (LV) is a detrimental
complication to acute myocardial infarction (MI) charac-
terized by increased chamber volume, altered chamber
geometry, and progressive deterioration of ventricular
function.1 LV remodeling is directly related to poor sur-
vival rate2 and has been speculated to be one of the
main reasons for the high and still increasing prevalence
From the aDanish Research Center of Magnetic Resonance, the bDepartment of Car-
diology, Copenhagen University Hospital, Hvidovre, cIntegrated Health Care Solu-
tions, F. Hoffmann-La Roche Ltd, and the dDepartment of Cardiology and
Endocrinology, Copenhagen University Hospital, Frederiksberg.
Supported by The Danish Heart Foundation (grants no. 97-1-3-62-22491, 97-2-4-
56-22548, 98-1-4-55-22598, and 99-1-3-62-22698), the Research Fund of the
Copenhagen Hospital Corporation, and the Novo Nordisk Foundation. Analysis for
N-terminal pro brain natriuretic peptide was provided by F. Hoffmann-La Roche Ltd,
Basel, Switzerland.
Submitted April 6, 2001; accepted September 18, 2001.
Reprint requests: Dr Jens C. Nilsson, Danish Research Center of Magnetic Reso-
nance, Section 340, Copenhagen University Hospital Hvidovre, Kettegaard Allé
30, DK-2650 Hvidovre, Denmark. E-mail: jenschrn@magnet.drcmr.dk
Copyright 2002, Mosby, Inc. All rights reserved.
0002-8703/2002/$35.00 + 0 4/1/120293
doi:10.1067/mhj.2002.120293
of heart failure in Western countries.3 These serious
aspects have stimulated great interest in antiremodeling
therapy. In current clinical practice, reperfusion ther-
apy4,5 and angiotensin-converting enzyme (ACE)
inhibitors6,7 seem to account for the most important
antiremodeling actions. In addition, β-blockers8 have
been shown to attenuate chronic LV dilatation, but their
effects in the aftermath of MI are controversial.9
The existence of effective antiremodeling therapy
underscores the importance of early identification and
targeting of individuals at high risk. Lately, interest has
focused on simple biochemical markers for LV remodel-
ing, and particularly, brain natriuretic peptide (BNP)
and its amino-terminal portion N-terminal pro BNP (NT-
proBNP) appear promising. Both are released from ven-
tricular myocardium in response to increased wall
stress,10 which is known to be one of the major forces
driving LV remodeling.1,11 This may imply that individu-
als in high risk of remodeling could be identified by ele-
vated levels of these peptides before chamber dilatation
actually occurrs. This possibility is supported by recent
American Heart Journal
Volume 143, Number 4
studies in which baseline levels of BNP and NT-proBNP
have been shown to reflect late LV systolic function
after MI.12,13
The purpose of this work was to study the frequency
and extent of LV dilatation in a modern infarct popula-
tion who was offered antiremodeling treatment in com-
pliance with daily clinical practice. Furthermore, we
wanted to explore the value of NT-proBNP as a baseline
marker for later LV dilatation.
Methods
Patients
Fifty patients with a first and transmural MI were recruited
from the Department of Cardiology, Copenhagen University
Hospital Hvidovre, from October 1997 to April 1999.
Inclusion criteria included written informed consent and a
diagnosis of MI on the basis of the presence of 3 conditions:
an episode of chest pain of 30 minutes or more duration resis-
tant to nitroglycerin treatment; electrocardiographic ST-seg-
ment elevation 2 mm or more in at least 2 contiguous precor-
dial leads or 1 mm or more in at least 2 contiguous extremity
leads; peak creatine kinase (CK) level ≥400 units/L (≥2× the
upper healthy limit), with the fraction of CK-MB being ≥5%.
Exclusion criteria with respect to magnetic resonance imaging
(MRI) were claustrophobia, atrial fibrillation, and implanted
ferromagnetic devices. In addition, patients with significant
valvular heart disease were excluded.
CK level was measured on admission and after 4, 8, 12, 24,
and 48 hours. Electrocardiograms were obtained on admis-
sion, before thrombolytic treatment, 1 hour after thrombolytic
treatment, and once daily after treatment. Reperfusion was
determined as a ≥20% reduction in the sum of ST-segment ele-
vation in the electrocardiogram recorded after thrombolytic
treatment, as compared with the electrocardiogram recorded
before this treatment.14 Before discharge, patients underwent
bicycle exercise testing—anginal pain, decreasing blood pres-
sure (BP), or ST-segment depression were taken as signs of
ischemia. Patients <80 years of age who had ischemia develop
at exercise testing were referred to coronary angiography,
from which possible revascularization intervention was
decided. Patients who underwent percutaneous transluminal
coronary angioplasty or coronary artery bypass grafting con-
tinued in the study.
In addition, 20 healthy controls with no history or symp-
toms of heart disease or other chronic disease underwent
examination once to achieve reference intervals for LV vol-
umes and NT-proBNP. The study was approved by the
regional ethics committee, and all participating subjects sup-
plied written informed consent.
Magnetic resonance imaging
Each patient was scheduled for examination with MRI at 1
week, 1 month, 3 months, 6 months, and 1 year after the
ischemic insult. The study was conducted with a 1.0-T whole
body scanner (Impact Expert Magnetom, Siemens AG, Erlan-
gen, Germany) equipped with a standard phased array chest
coil. Each slice of the LV was acquired during 15 heartbeats
with an electrocardiogram-gated cinematographic breath-hold
pulse sequence.15 Field of view was 263 × 350 mm2, matrix
Nilsson et al 697
was 126 × 256, slice thickness was 10 mm, interslice gap was
0, and temporal resolution was 55 ms. The slices were posi-
tioned in the true short axis of the LV consecutively from the
apex towards the basis of the heart. Typically, the ventricle
was covered with 9 to 11 slices. During the MRI examina-
tions, patients were monitored with electrocardiography con-
tinuously. The average heart rate (HR) during the cinemato-
graphic imaging was used in the data analysis. BP was
measured after 20 minutes rest in a supine position. Height
was measured at the first examination, whereas weight was
measured at every visit. Height and weight were used in the
calculation of body surface area.16
All examinations were analyzed in 1 batch and in random
order by 1 observer in a blinded manner. The end-diastolic
frame of each cinematographic slice was defined as the frame
acquired corresponding to the R-wave in the electrocardio-
gram. The end-systolic frame was identified as the frame with
the smallest total LV lumen. In each examination, the number
of slices to include in end-diastole and end-systole were
decided according to predefined criteria.17 Each image was
analyzed with manual tracing. LV end-diastolic volume index
(LVEDVI) and LV end-systolic volume index (LVESVI) were
calculated by addition of the planimetrically measured areas of
the cavities multiplied by slice thickness. LV myocardial mass
index (LVMI) was determined in a similar way with a density
factor of 1.05 g/cm3, including the papillary muscles as part of
the myocardium. LV stroke volume index (LVSVI) was calcu-
lated by subtraction of LV end-systolic volume from end-dias-
tolic volume. LV ejection fraction (LVEF) was calculated with
division of LVSVI with LVEDVI. Cardiac output was calculated
by multiplication of LVSVI by HR. All variables, except LVEF,
were indexed by division with body surface area.
NT-proBNP
Blood samples were taken immediately before the MRI
examination after 30 minutes rest in a sitting position. To
exclude the effects of a potential hormonal circadian rhythm,
all subjects were examined between 11 AM and 3 PM. Analysis
for NT-proBNP was performed by LabConsult GmbH, Denzlin-
gen, Germany, with a novel immunoassay on the basis of a
sandwich format with unextracted ethylenediamine
tetraacetic acid plasma. The sensitivity of the assay was <3.0
pmol/L, and the intraassay and interassay coefficients of varia-
tion were 1.3% and 4.8%, respectively.18
Statistics
Dichotomous variables were compared with Fisher’s exact
test, and trend over groups was analyzed with χ2 test for
trend. Continuous variables were compared with Student
unpaired t test or with analysis of variance. Normal distribu-
tion of data was examined with histograms and normal plots.
Peak CK level, NT-proBNP, HR, BP, LVEDVI, LVESVI, and
LVMI were normally distributed after logarithmic
transformation. Intercategory comparisons regarding LV mea-
sures and NT-proBNP were performed on 1-week and 1-year
data only. The linear associations between baseline NT-
proBNP and LV measures were evaluated with Pearson corre-
lation analysis. Serial measurements were analyzed with sto-
chastic regression analysis. Model control results confirmed
that all conditions for the statistical procedure had been ful-
 American Heart Journal
698 Nilsson et al April 2002

Table I. Patient characteristics

All Non-dilators Transient dilators Dilators
(n = 42) (n = 18) (n = 12) (n = 12)

Age (y) 58 (11.7) 58 (13.1) 60 (11.9) 54 (9.4)

Male sex (%) 35 (83) 16 (89) 11 (92) 8 (67)
Hypertension (%) 10 (24) 5 (28) 2 (17) 3 (25)
Diabetes (%) 5 (12) 2 (11) 2 (17) 1 (8)
Infarct location (%)
Anterior 34 (81) 13 (72) 9 (75) 12 (100)
Inferior 8 (19) 5 (28) 3 (25) 0 (0)
Peak creatine kinase (arbitrary units/L) 1590 (1213–2085) 1293 (865–1934) 1151 (804–1646) 2944 (1624–5336)*
Thrombolytic treatment (%) 35 (83) 17 (94) 9 (75) 9 (75)
Thrombolytic delay (hours) 3.2 (2.4–4.3) 2.9 (1.8–4.7) 4.6 (2.6–8.1) 2.8 (1.5–5.2)
Reperfusion 84% 93% 88% 67%
Ischemia at exercise test (%) 17/39 (44) 4/17 (24) 6/12 (50) 7/10 (70)†
Medication at discharge (%)
ACE inhibitors 17 (40) 8 (44) 3 (25) 6 (50)
β-antagonists 30 (71) 13 (72) 9 (75) 8 (67)
Nitrates 5 (12) 0 (0) 1 (8) 4 (33)‡
Revascularization within study period (%)
PTCA 12 (29) 6 (33) 3 (25) 3 (25)
CABG 6 (14) 2 (11) 2 (17) 2 (17)

Age data presented as mean (standard deviation). Peak creatine kinase data presented as geometric mean (95% CI). Thrombolytic treatment was either streptokinase or
alteplase. Thrombolytic delay (time from appearance of symptoms until beginning of trombolytic treatment) data presented as geometric mean (95% CI). Reperfusion data pre-
sented as percentage of patients with reperfusion determined with electrocardiogram criteria. Ischemia at exercise test data presented as number of patients with ischemia at
exercise test/number of patients with conclusive test results. PTCA, Percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass grafting.
*P = .02 and P = .008 for dilators versus non-dilators and transient dilators, respectively.
†P = .02 for trend over categories.
‡P = .02 for dilators versus pooled data from non-dilators and transient dilators.

filled. Multiple logistic regression analysis (backward selec-
tion) was used to analyze the value of different baseline char-
acteristics as independent predictors of the dichotomous out-
come of being a dilator or not. Finally, the diagnostic abilities
of NT-proBNP were evaluated with receiver operating charac-
teristic (ROC) curve analysis. All tests were 2-sided, and a sig-
nificance level of 5% was used. Data analysis was performed
in the Statistical Analysis System version 6.12 (SAS Institute
Inc, Cary, NC).
Results
Of the 50 patients initially included, 8 patients were
excluded after enrollment. Three patients were
excluded because of death, reinfarction, or inability to
cooperate with the MRI examination. One patient who
had extreme LV dimensions was excluded because of
anabolic steroid abuse. Four patients were excluded
because 1-year examinations were not performed; none
of these patients were lost to follow-up examination
because of aggravation of disease. Thus, 42 patients
were left for the analysis. Patient characteristics are
listed in Table I.
LV volumes, function, and mass
In the entire patient sample, LVEDVI, LVESVI, LVSVI,
LVEF, and LVMI remained constant during the 1-year
follow-up period. To analyze data at the individual
level, all patients were categorized according to their
specific 1-year change in LVEDV, because this is consid-
ered a fundamental measure of remodeling.11 These
volume changes were related to the repeatability coeffi-
cient19 for LVEDV as measured with MRI, which has
been determined by Lorenz et al17 to be approximately
11.8 mL.17 Accordingly, patients with a change in
LVEDV of ≤11.8 mL were categorized as non-dilators (n
= 18). Patients with an increase in LVEDV of >11.8 mL
were categorized as dilators (n = 12), and patients with
a decrease in LVEDV of >11.8 mL were categorized as
transient dilators (n = 12), because these patients had
LV dilatation at baseline.
In non-dilators, LVEDVI, LVESVI, and LVSVI remained
constant during the follow-up period, whereas the tran-
sient dilators had decreases of 19%, 23%, and 13%,
respectively, and at the final examination, non-dilators
and transient dilators were similar regarding these vari-
ables. In dilators, LVEDVI, LVESVI, and LVSVI increased
by 24%, 22%, and 27%, respectively, during the follow-
up period, and after 1 year, both LVEDVI and LVESVI,
but not LVSVI, were significantly different from values
in non-dilators and transient dilators (Figure 1). As com-
pared with healthy values, LVEDVI was elevated in dila-
tors at the final examination, whereas LVESVI was ele-
vated in all 3 patient categories after both 1 week and 1
year.
LVEF at baseline was higher in non-dilators and tran-
sient dilators as compared with dilators (mean [stan-
American Heart Journal
Volume 143, Number 4 Nilsson et al 699

Figure 1

Course of LVEDVI, LVESVI, LVSVI, and LVMI over time in dilators (▲), transient dilators (◊), and non-dilators (
). Error bars indicate
95% CI for geometric mean (LVEDVI, LVESVI, and LVMI) and mean (LVSVI). Shaded rectangles mark reference interval for mean
obtained from healthy controls. →, ↓, and ↑ indicate no change, decrease, and increase, respectively, over time. P values are from
stochastic regression analyses. Intercategory comparisons at 1 week and 1 year are listed over examination times.

dard deviation], 59.8% [11.7%] and 57.8% [9.1%] vs
47.4% [11.3%]; P = .008 and P = .02) and remained
unchanged during the follow-up period in all 3 patient
categories. LVEF in all 3 patient categories was signifi-
cantly less than values in the healthy controls (68.8%
[5.6%]). Cardiac index at baseline was lower in dilators
compared with transient dilators (2.46 [0.57] vs 3.12
[0.38] L • m–2 • min–1; P = .003) but decreased over
time in the latter (P = .006), and at the final examina-
tion, there were no intercategory differences. LVMI was
elevated throughout the study period in the dilators,
whereas the transient dilators had a 7% decrease from
an elevated level at 1 week to a healthy level after 1
year (Figure 1).
HR and blood pressure
HR was the same in all 3 categories after 1 week and
1 year but decreased by 14% over time in the dilators (P
= .01). Systolic BP was lower in non-dilators as com-
pared with transient dilators at baseline (geometric
mean [95% CI] 112 mm Hg [106-119 mm Hg] vs 125
mm Hg [115-136 mm Hg]; P = .03). Systolic BP
increased by 13% in dilators (P = .002), and both sys-
tolic and diastolic BP increased by 9% and 6% (P = .01
and P = .006) over time in non-dilators.
Predictors of LV dilatation
Peak CK level was significantly higher in the dilators as
compared with the 2 other patient categories, and there
was a trend towards a decreasing frequency of reperfu-
sion from non-dilators over transient dilators to dilators
(P = .11; Table I). In addition, the frequency of ischemia
at exercise testing increased significantly from the non-
dilators over the transient dilators to the dilators (Table
I). Furthermore, there were trends towards a higher pro-
portion of female patients and anterior infarct location
among dilators as compared with pooled data from the 2
other categories (P = .09 and P = .08).
Although decreasing over time, NT-proBNP level was
higher than control values in all 3 patient categories

700 Nilsson et al
Figure 2
Course of NT-proBNP over time in dilators (▲), transient dila-
tors (◊), and non-dilators (
). Error bars indicate 95% CI for
geometric mean. Shaded rectangle marks reference interval
for geometric mean obtained from healthy controls. ↓ indicates
decrease over time. P values are from stochastic regression
analyses. Intercategory comparisons at 1 week and 1 year are
listed over examination times.
throughout the study period, except for the final mea-
surement in the non-dilators (Figure 2). NT-proBNP level
was significantly higher in dilators than in non-dilators
after 1 week and 1 year. NT-proBNP level at baseline cor-
related with LVEDVI, LVESVI, LVEF, and LVMI at baseline
(r = 0.33, P = .06; r = 0.41, P = .02; r = –0.40, P = .02; r =
0.41, P = .01). Furthermore, and more important, NT-
proBNP level at baseline was significantly related to
LVEDVI and LVESVI (r = 0.55, P = .0007; r = 0.41, P =
.02) after 1 year but not to LVEF or LVMI. In addition,
NT-proBNP level at baseline was related to the 1-year
change in LVEDVI (r = 0.42, P = .01; Figure 3).
Finally, multivariate analyses were performed to evalu-
ate a number of clinical variables as predictors of later LV
dilatation. When age, gender, infarct location, peak CK
level, and LVEF were entered into the analysis, female
gender, peak CK level, and LVEF were identified as inde-
pendent predictors of LV dilatation. However, when NT-
proBNP level was added to the model, only low age and
NT-proBNP level came out as independent markers for
LV dilatation (Table II). ROC analysis results revealed that
a baseline NT-proBNP level of more than 115 pmol/L
identified the dilators with a sensitivity and specificity of
89% and 68% and positive and negative predictive values
of 50% and 94% (area under the curve = 0.77).
Discussion
Main morphologic findings
In this study, LV volumes, myocardial mass, and systolic
function were unchanged over time in the entire undi-
American Heart Journal
April 2002
Figure 3
Relation between NT-proBNP at baseline and change in
LVEDVI from 1 week to 1 year (regression line and 95% confi-
dence limits).
vided patient sample. However, when results were evalu-
ated at the level of the individual patients, it turned out
that approximately 30% had increases in LVEDVI and
LVESVI, approximately 30% had decreases in LVEDVI and
LVESVI, and the remaining 40% had unchanged LV vol-
umes during the study period. LVEF was significantly
reduced as compared with healthy values but was
unchanged over time in all 3 patient categories.
As expected, the magnitude of LV dilatation in this
study appeared lower than that reported in studies from
the prethrombolytic era.20,21 In contrast, there seemed to
be no clear differences between these results and those
of studies in which most patients received thrombolytics
but none or almost none were given ACE inhibitors.22,23
Although there may be many reasons for this surprising
finding, one obvious explanation could be that only
about 50% of the dilators in this study received an ACE
inhibitor during the study period. LVEF was only moder-
ately depressed among the dilators. Therefore, it seems
that patients who will develop LV dilatation later will
likely be missed if identification of patients at high risk is
made solely on the basis of LV systolic impairment or
symptoms of heart failure. This possibility emphasizes the
need for alternative high-risk measures.
Prediction of LV dilatation
In recent years, an increasing amount of observations
have pointed to BNP and NT-proBNP as high-risk mark-
ers after MI. In continuation of these findings, this
study provides direct evidence for a relation between
NT-proBNP measured at discharge and subsequent LV
remodeling.
Plasma levels of NT-proBNP were significantly higher
in the dilators compared with the non-dilators at dis-
charge (Figure 2) and correlated with LVEDVI and
American Heart Journal
Volume 143, Number 4
Table II. Predictors of left ventricular dilatation
Multivariate analysis (n = 41)
Regression coefficient (SE) P value
Age – –
Sex –3.46 (1.57) .03
Infarct location – –
log10–CK 2.74 (1.02) .007
LVEF –12.79 (4.51) .005
log10–NT-proBNP – –
Sex, female = 0, male = 1; Infarct location was anterior or inferior. LVEF at baseline was measured with magnetic resonance imaging. log10-CK is logarithm to base 10 of peak
CK (arbitrary units/L). log10–NT-proBNP is logarithm to base 10 of NT-proBNP at baseline (pmol/L).
LVESVI at the final examination. Moreover, the NT-
proBNP level at discharge correlated with the 1-year
change in LVEDVI (Figure 3), which is in agreement
with previous results on BNP from somewhat smaller
studies with shorter follow-up periods.24,25 Even more
interestingly, multivariate analysis results revealed that
the NT-proBNP level at discharge eliminated both peak
CK level and LVEF (measured with MRI) as predictors of
subsequent LV dilatation (Table II). This seems to indi-
cate that NT-proBNP holds predictive information
beyond the magnitude of ischemic damage, which is
readily explainable, because NT-proBNP is released from
ventricular myocardium in response to increased wall
stress (known to be one of the major forces driving LV
remodeling). In other words, NT-proBNP may directly
reflect the impetus for LV remodeling. Results from this
study on the diagnostic abilities of NT-proBNP support
this explanation because ROC analysis results revealed
that baseline levels of NT-proBNP accurately identified
patients in whom LV dilatation later developed.
With the consideration that heart failure is the clinical
manifestation of LV remodeling in advanced stages,
these findings seem similar to results from previous
studies in which BNP or NT-proBNP were identified as
independent predictors of postinfarction heart failure
or death.12,13,26 However, in contrast to some of these
studies,12,13 we failed to show any relation between
NT-proBNP level measured at baseline and late LVEF.
This may have been a consequence of the longer fol-
low-up period in this study and perhaps also of a less
than optimal time-point of NT-proBNP sampling.27
Additional findings
The slow decline in the plasma levels of NT-proBNP
over time in all 3 patient categories (Figure 2) suggests
that myocardial wall stress may have decreased during
the follow-up period. According to the modified Laplace
relationship, a decrease in global wall stress implies a
decrease in the ratio between chamber radius and wall
Nilsson et al 701
Multivariate analysis (n = 34)
Regression coefficient (SE) P value
–0.133 (0.04) .003
– –
– –
– –
– –
3.15 (1.16) .007
thickness over time. This is, however, not supported by
the current data (Figure 1), so mechanisms other than a
decrease in global wall stress must be active.
The time course of LV volumes in the transient dilators
is somewhat intriguing. From an elevated baseline level,
LVEDVI and LVESVI decreased gradually and reached the
level of the non-dilators after weeks to months. We
found no obvious explanation for the differing time
course in these 2 patient categories; however, other
authors have reported similar findings.23,28,29
The large dispersion of LVSVI among the 3 patient
categories at baseline, which became more uniform
during the follow-up period (Figure 1), may point to
LVSVI as an essential physiologic variable being regu-
lated to healthy or near-healthy values with adaptive
processes. In an LV characterized by a reduced LVSVI,
this adaptation, as previously proposed,1 seems to
involve an increase in LVEDVI.
Limitations of this study
Baseline examinations were performed 1 week after
the MI when infarct expansion had possibly already
taken place, leaving the starting point of LV measures
unknown. Furthermore, the limited number of study
participants calls for caution in making conclusions,
particularly regarding negative findings.
Conclusion
In this 1-year follow-up study of patients with a first
transmural MI, approximately 30% had significant incre-
ments develop in LVEDVI and LVESVI. LVEF remained
unchanged. Patients in whom LV dilatation developed
could be identified early after the MI with elevated
plasma levels of NT-proBNP.
We thank the laboratory technicians and radiogra-
phers at the Danish Research Center of Magnetic Res-
onance, H:S Hvidovre Hospital.

702 Nilsson et al
References
1. Pfeffer MA, Braunwald E. Ventricular remodeling after myocardial
infarction. Experimental observations and clinical implications. Cir-
culation 1990;81:1161-72.
2. White HD, Norris RM, Brown MA, et al. Left ventricular end-systolic
volume as the major determinant of survival after recovery from
myocardial infarction. Circulation 1987;76:44-51.
3. Coats AJ. Is preventive medicine responsible for the increasing
prevalence of heart failure? Lancet 1998;352(Suppl 1):39-41.
4. Marino P, Zanolla L, Zardini P. Effect of streptokinase on left ventric-
ular modeling and function after myocardial infarction: the GISSI
(Gruppo Italiano per lo Studio della Streptochinasi nell’Infarto Mio-
cardico) trial. J Am Coll Cardiol 1989;14:1149-58.
5. Popovic AD, Neskovic AN, Marinkovic J, et al. Acute and long-term
effects of thrombolysis after anterior wall acute myocardial infarc-
tion with serial assessment of infarct expansion and late ventricular
remodeling. Am J Cardiol 1996;77:446-50.
6. Bonarjee VV, Carstensen S, Caidahl K, et al. Attenuation of left ven-
tricular dilatation after acute myocardial infarction by early initiation
of enalapril therapy. CONSENSUS II Multi-Echo Study Group. Am
J Cardiol 1993;72:1004-9.
7. St John-Sutton M, Pfeffer MA, Plappert T, et al. Quantitative two-
dimensional echocardiographic measurements are major predictors
of adverse cardiovascular events after acute myocardial infarction.
The protective effects of captopril. Circulation 1994;89:68-75.
8. Groenning BA, Nilsson JC, Sondergaard L, et al. Antiremodeling
effects on the left ventricle during beta-blockade with metoprolol in
the treatment of chronic heart failure. J Am Coll Cardiol 2000;36:
2072-80.
9. Yoshitomi Y, Kojima S, Yano M, et al. Long-term effects of bisoprolol
compared with imidapril on left ventricular remodeling after reperfu-
sion in acute myocardial infarction: an angiographic study in pa-
tients with maintained vessel patency. Am Heart J 2000;140:E27.
10. Magga J, Vuolteenaho O, Tokola H, et al. B-type natriuretic peptide:
a myocyte-specific marker for characterizing load-induced alterations
in cardiac gene expression. Ann Med 1998;30(suppl 1):39-45.
11. Cohn JN, Ferrari R, Sharpe N. Cardiac remodeling—concepts and
clinical implications: a consensus paper from an international forum
on cardiac remodeling. Behalf of an International Forum on Car-
diac Remodeling. J Am Coll Cardiol 2000;35:569-82.
12. Richards AM, Nicholls MG, Yandle TG, et al. Plasma N-terminal
pro-brain natriuretic peptide and adrenomedullin: new neurohor-
monal predictors of left ventricular function and prognosis after
myocardial infarction. Circulation 1998;97:1921-9.
13. Richards AM, Nicholls MG, Yandle TG, et al. Neuroendocrine pre-
diction of left ventricular function and heart failure after acute
myocardial infarction. The Christchurch Cardioendocrine Research
Group. Heart 1999;81:114-20.
14. Clemmensen P, Ohman EM, Sevilla DC, et al. Changes in standard
electrocardiographic ST-segment elevation predictive of successful
reperfusion in acute myocardial infarction. Am J Cardiol 1990;66:
1407-11.
15. Sakuma H, Fujita N, Foo TK, et al. Evaluation of left ventricular vol-
American Heart Journal
April 2002
ume and mass with breath-hold cine MR imaging. Radiology 1993;
188:377-80.
16. Guyton AC. Energetics and metabolic rate. In: Guyton AC, ed.
Medical physiology. 7th ed. Philadelphia: WB Saunders; 1986. p.
841-8.
17. Lorenz CH, Walker ES, Morgan VL, et al. Normal human right and
left ventricular mass, systolic function, and gender differences by
cine magnetic resonance imaging. J Cardiovasc Magn Reson
1999;1:7-21.
18. Karl J, Borgya A, Gallusser A, et al. Development of a novel, N-ter-
minal-proBNP (NT-proBNP) assay with a low detection limit. Scand
J Clin Lab Invest Suppl 1999;230:177-81.
19. Bland JM, Altman DG. Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet 1986;1:307-
10.
20. Gadsboll N, Hoilund-Carlsen PF, Badsberg JH, et al. Late ventricu-
lar dilatation in survivors of acute myocardial infarction. Am J Car-
diol 1989;64:961-6.
21. Jeremy RW, Allman KC, Bautovitch G, et al. Patterns of left ventricu-
lar dilation during the six months after myocardial infarction. J Am
Coll Cardiol 1989;13:304-10.
22. Gaudron P, Eilles C, Kugler I, et al. Progressive left ventricular dys-
function and remodeling after myocardial infarction. Potential mech-
anisms and early predictors. Circulation 1993;87:755-63.
23. Walsh JT, Batin PD, Hawkins M, et al. Ventricular dilatation in the
absence of ACE inhibitors: influence of haemodynamic and neuro-
hormonal variables following myocardial infarction. Heart 1999;
81:33-9.
24. Nagaya N, Nishikimi T, Goto Y, et al. Plasma brain natriuretic pep-
tide is a biochemical marker for the prediction of progressive ven-
tricular remodeling after acute myocardial infarction. Am Heart J
1998;135:21-8.
25. Nagaya N, Goto Y, Nishikimi T, et al. Sustained elevation of
plasma brain natriuretic peptide levels associated with progressive
ventricular remodelling after acute myocardial infarction. Clin Sci
(Colch) 1999;96:129-36.
26. Omland T, Aakvaag A, Bonarjee VV, et al. Plasma brain natriuretic
peptide as an indicator of left ventricular systolic function and long-
term survival after acute myocardial infarction. Comparison with
plasma atrial natriuretic peptide and N-terminal proatrial natriuretic
peptide. Circulation 1996;93:1963-9.
27. Talwar S, Squire IB, Downie PF, et al. Profile of plasma N-terminal
proBNP following acute myocardial infarction. Correlation with left
ventricular systolic dysfunction. Eur Heart J 2000;21:1514-21.
28. Giannuzzi P, Temporelli PL, Bosimini E, et al. Heterogeneity of left
ventricular remodeling after acute myocardial infarction: results of
the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto
Miocardico-3 Echo Substudy. Am Heart J 2001;141:131-8.
29. Aikawa Y, Rohde L, Plehn J, et al. Regional wall stress predicts ven-
tricular remodeling after anteroseptal myocardial infarction in the
Healing and Early Afterload Reducing Trial (HEART): an echocar-
diography-based structural analysis. Am Heart J 2001;141:234-
42.