Clinical Investigation

Adverse diastolic remodeling after reperfused

ST-elevation myocardial infarction:
An important prognostic indicator
Tuan L. Nguyen, MBBS, a,c Justin Phan, MBBS, a,c Jarred Hogan, BMed, c Leia Hee, BMedSci, MClinTRes, PhD, a
Daniel Moses, BSc, MBBS, MEngSc, b,c James Otton, MBBS, PhD, a,c Upul Premawardhana, MBBS, MMed, a,c
Rohan Rajaratnam, MBBS, a,c Craig P. Juergens, MBBS, DMedSc, a,c Hany Dimitri, MBBS, PhD, a,c
John K. French, BMedSc, MSc, MBChB, PhD, a,c David Richards, BSc (Med), MBBS, MD, a,c and
Liza Thomas, MBBS, PhD a,c NSW, Australia

Objectives We sought to determine the relationship of adverse diastolic remodeling (ie, worsening diastolic or persistent
restrictive filling) with infarct scar characteristics, and to evaluate its prognostic value after ST-segment elevation myocardial
infarction (STEMI).
Background Severe diastolic dysfunction (restrictive filling) has known prognostic value post STEMI. However, ongoing
left ventricular (LV) remodeling post STEMI may alter diastolic function even if less severe.
Methods and results There were 218 prospectively recruited STEMI patients with serial echocardiograms
(transthoracic echocardiography) and cardiac magnetic resonance imaging (CMR) performed, at a median of 4 days (early)
and 55 days (follow-up). LV ejection fraction and infarct characteristics were assessed by CMR, and comprehensive diastolic
function assessment including a diastolic grade was evaluated on transthoracic echocardiography. ‘Adverse diastolic
remodeling’ occurred if diastolic function grade either worsened (≥1 grade) between early and follow-up imaging, or
remained as persistent restrictive filling at follow-up. Follow-up infarct scar size (IS) predicted adverse diastolic remodeling
(area under the curve 0.86) and persistent restrictive filling (area under the curve 0.89). The primary endpoint of major adverse
cardiovascular events (MACE) occurred in 48 patients during follow-up (mean, 710 ± 79 days). Kaplan-Meier analysis
showed that adverse diastolic remodeling (n = 50) and persistent restrictive filling alone (n = 33) were significant predictors of
MACE (both P b .001). Multivariate Cox analysis, when adjusted for TIMI risk score and CMR IS, microvascular obstruction,
and LV ejection fraction, showed adverse diastolic remodeling (HR 3.79, P b .001) was an independent predictor of MACE,
as was persistent restrictive filling alone (HR 2.61, P = .019).
Conclusions Larger IS is associated with adverse diastolic remodeling. Following STEMI, adverse diastolic remodeling
is a powerful prognostic marker, and identifies a larger group of ‘at-risk’ patients, than does persistent restrictive filling alone.
(Am Heart J 2016;180:117-27.)

Left ventricular (LV) infarct scar size (IS), microvascu-

a
lar obstruction (MVO) and reduced LV function on CMR,
From the Cardiology Department, Liverpool Hospital, Sydney, NSW, Australia,
b
Radiology Department, Liverpool Hospital, Sydney, NSW, Australia, and cSouth Western
are established predictors of adverse cardiovascular
Sydney Clinical School, The University of NSW, Sydney, NSW, Australia. outcomes and mortality, after acute ST-segment elevation
Sources of funding/Disclosures: TLN was supported by a National Health and Medical myocardial infarction (STEMI). 1,2 In addition, diastolic
Research Council and National Heart Foundation postgraduate scholarship. The
Cardiology department has received unrestricted educational grants from Biotronik,
dysfunction has demonstrated prognostic value in STEMI
Medtronic and St Jude Medical. patients treated with thrombolytic therapy. 3,4 Diastolic
Conflict of interest: All other authors have no disclosures relevant to the contents of this function is altered by myocardial necrosis and microvas-
manuscript. The authors are solely responsible for the design and conduct of this study all
cular dysfunction consequent to STEMI. In addition,
study analyses, the drafting and editing of the manuscript, and its final contents.
Submitted May 9, 2016; accepted May 9, 2016.
myocardial edema and infarct scar formation can lead to
Reprint requests: Liza Thomas, MBBS, PhD, Cardiology Department, Liverpool Hospital, increased LV wall stiffness and altered LV filling.
Elizabeth Street, Liverpool, NSW, 2170, Sydney, Australia. There are established echocardiographic parameters to
E-mail: nguyenletuan@hotmail.com
determine diastolic function (including transmitral pulse
0002-8703
Crown Copyright © 2016 Published by Elsevier Inc. All rights reserved. wave Doppler and tissue Doppler assessment). The
http://dx.doi.org/10.1016/j.ahj.2016.05.020 diastolic function grade incorporates several parameters,

118 Nguyen et al
Table I. Diastolic function grade assessment
Normal (Grade 0)
Mitral inflow PW Doppler E/A = 0.8–1.5
DT = 160-200 ms
Tissue Doppler⁎ E/e′ ≤ 8
Pulmonary inflow PW Doppler S=D
⁎ Mean of septal and lateral tissue Doppler measurements. Adapted from Moller et al5 and Nagueh et al8 DT, Deceleration time; PW, pulse wave.
to define severity of diastolic dysfunction (Table I). 5
Severe diastolic dysfunction, or restrictive filling, is a
powerful prognostic marker that identifies high-risk
patients post STEMI; however it may only identify a
limited proportion of at-risk patients. Diastolic function is
dynamic, and diastolic grade can be altered by LV
remodeling, IS and myocardial viability, all of which are
important factors post STEMI. 6 However, there is a
paucity of literature on diastolic function ‘remodeling’
and in particular, its prognostic value post STEMI.
Furthermore, evaluation of the relationship between
diastolic function and IS has not been well described.
Our study aims were to determine the relationship of
adverse diastolic remodeling with CMR infarct scar
characteristics, and its prognostic significance after
acute STEMI treated by primary percutaneous coronary
intervention (PCI) or thrombolytic reperfusion.
Methods
Study population
We prospectively evaluated patients presenting with
acute STEMI (defined as clinical symptoms with
ST-segment elevation in 2 or more contiguous leads),
treated by primary PCI, reperfusion by thrombolysis with
nonemergent PCI, or rescue PCI after unsuccessful
thrombolysis, who had presented to our tertiary care
facility. The study population included both patients with
first presentation STEMI, which have been previously
described, 7 and patients with a prior history of myocar-
dial infarction. Electrocardiograms and/or coronary
angiography reports were used to confirm a prior history
of myocardial infarction. The exclusion criteria included:
severe chronic kidney disease (eGFR b30 mL/min per
1.73 m 2, or renal replacement therapy), prior valvular/
coronary bypass surgery or congenital heart disease,
known cardiomyopathy, previous history of atrial fibril-
lation, coexistent conditions with survival of b1 year or
significant psychiatric illness, CMR exclusions (including
claustrophobia, gadolinium allergy and ferrous metallic
implants), and age b18 years or N85 years. PCI was
considered successful if there was TIMI III flow and b20%
residual stenosis of the culprit vessel was achieved, with
resolution of their presenting symptoms as determined by
the interventionalist. Detailed patient demographics,
discharge medications, and cardiac risk factors were
recorded after obtaining patient consent. Patients under-
American Heart Journal
October 2016
Impaired relaxation (Grade 1) Pseudonormal (Grade 2) Restrictive filling (Grade 3)
E/A b 0.8 E/A = 0.8–1.5 E/A N 1.5
DT N200 ms DT = 160-200 ms DT b160 ms
E/e′ ≤ 8 E/e′ = 9–12 E/e′ ≥ 13
SND SbD SNND
went ‘early’ (median 4 days, interquartile range 3-7 days)
and ‘follow-up’ (median 55 days, interquartile range 46-64
days) transthoracic echocardiogram (TTE) and CMR, after
STEMI presentation. Patients were followed up to
determine adverse clinical events for up to 2 years (730
days) post STEMI.
Diastolic function assessment
Comprehensive TTEs were performed on commercial-
ly available Vivid E9 machines (GE Healthcare, Norway).
All measurements and analysis were performed offline,
using an EchoPAC Clinical Workstation (GE Healthcare,
Version 12). Pulsed Doppler mitral inflow measurements
were obtained from the apical four-chamber view, with
the sample volume placed at the tips of the mitral valve
leaflets. Early mitral inflow velocity (E) late mitral inflow
velocity (A), E/A ratio, and mitral E wave deceleration
time were measured. Pulmonary venous inflow measure-
ments were acquired by placing the sample volume 1–2 cm
into the superior pulmonary vein. Tissue Doppler early
diastolic mitral annular velocity (e′) was an average of septal
and lateral e′. The E/e′ ratio was calculated using the average
e′ velocity. Three measurements were made on consecutive
cardiac cycles, and their average recorded.
All individual diastolic parameters mentioned above were
evaluated using established criteria 5,8 to categorize patients
to a diastolic function grade (0 = normal, 1 = impaired
relaxation, 2 = pseudonormal, 3 = restrictive pattern filling)
with consensus between two blinded independent ob-
servers (Table I). ‘Adverse diastolic remodeling’ included
patients whose diastolic function grade increased (≥1 grade)
from the ‘early’ to ‘follow-up’ scans, as well as those with
persistent restrictive filling (grade 3) at ‘follow-up’.
CMR acquisition protocol
A detailed protocol with specific CMR parameters used,
has been previously described. 7 In brief, patients underwent
early and follow-up CMR, using a commercially available MRI
1.5 T scanner (Siemens Symphony, Germany), and a
standard CMR multisequence protocol with image se-
quences performed during breath-hold. 9 A 6-channel body
array coil and a spine coil were used. Cardiac synchroniza-
tion was obtained by retrospective vector electrocardio-
graphic gating. Cine images, using a steady state free
precession pulse sequence, were acquired in three
long-axis views, and contiguous short-axis images from LV
base to apex.
American Heart Journal
Volume 180
Post contrast late gadolinium enhancement CMR
images were obtained 8 to 10 minutes after bolus
injection of 0.10 mmol/kg gadoteric acid (Dotarem,
Guerbet, France), with ECG-triggered phase-sensitive
inversion recovery gradient echo sequence on consecu-
tive slices. The late gadolinium enhancement images
were obtained in the same planes as the cine sequences.
CMR analysis
Indexed LV volumes, LV mass, LV ejection fraction, and
LV scar characteristics (infarct size/mass, scar heteroge-
neity, and MVO) were quantified offline, using computer-
assisted manual planimetry of the endocardial and
epicardial contours on contiguous short axis images, by
commercially available software (CMR 42, version 4.1.8,
Circle Cardiovascular Imaging, Calgary, Canada). We
employed previously reported signal threshold versus
reference myocardium technique for the evaluation of
scar characteristics. 10,11 LV infarct scar area was defined
as the hyperenhanced area with signal intensity threshold
≥5 SD above a region of interest of normal ‘nulled’
myocardium. 9 Scar volume (cm 3 ) was derived by
multiplying scar area for each short axis image, by the
slice thickness, and scar mass (g) was obtained by
multiplying scar volume by the specific density of
myocardial muscle (1.05 g/ cm 3). IS was calculated as
the percentage of infarct mass of the total LV mass. MVO
was determined by manual contouring of areas of
hypoenhancement with surrounding hyperenhanced
myocardium, on the delayed enhancement inversion
recovery sequences. 11,12 If MVO was present, it was
expressed by percentage of total LV mass. IS and scar
characteristics were measured by 2 experienced readers,
who were blinded to clinical outcome events. Our
institution has previously reported excellent intra and
interobserver variability for CMR measurements. 7
Definitions of clinical endpoints
The primary study endpoint was major adverse
cardiovascular events (MACE), which was a composite
of all cause mortality, reinfarction, new or worsening
heart failure, cerebrovascular accident, and sustained
ventricular arrhythmias. For patients with more than 1
MACE, the primary endpoint was determined as time to
first MACE event. MACE events were obtained by patient
follow up in clinic or by telephone. MACE was also
corroborated by the patient's family physician, treating
cardiologist, or family. Documents/clinical charts were
reviewed on hospital database systems (if event occurred
within our hospital area network) or retrieval of clinical
documents (if event occurred within a different hospital
area health service) to confirm the event details.
Reinfarction criteria were based on the ESC/ACCF/
AHA/WHF consensus document of the third definition of
myocardial infarction. 13 A heart failure event occurred if
a patient required hospitalization for heart failure with
Nguyen et al 119
clinical or radiological evidence supportive of this
diagnosis, with response to diuretics and heart failure
therapy. For inpatients, this was defined as worsening
dyspnea with clinical or radiological evidence of pulmo-
nary edema requiring treatment with diuretics, balloon
pump or inotropic agents after enrolment. A cerebrovas-
cular accident was defined as a presentation with new
typical neurological deficits and symptoms N24 hours
(including hemiparesis, dysphasia, or other focal objec-
tive neurological signs), and radiological evidence of an
acute cerebrovascular accident (ischemic or hemorrhag-
ic). Sustained ventricular arrhythmias included aborted
sudden cardiac death (with documentation of ventricular
tachycardia/ventricular fibrillation) with successful resus-
citation, whilst for patients with implantable cardioverter
defibrillators, this was defined as a broad complex
tachycardia of ventricular origin with hemodynamic
instability or requiring device activated therapy. 14,15
Statistical analysis
Continuous data with normal distribution are presented
as mean and SD. Non-normally distributed variables were
reported as median with corresponding interquartile
range. Categorical variables are represented by frequen-
cies and percentages. Patients were stratified as those
with and without adverse diastolic remodeling. Compar-
isons between the groups were performed using Student
t test (normally distributed), Mann Whitney U test
(non-normally distributed) for continuous variables, and
Chi-square test for categorical variables. Correlations be-
tween parameters were examined by Pearson correlation for
continuous variables. Spearman rho correlation was used for
analysis of nonparametric data and ordinal variables.
Patients with adverse diastolic remodeling were also
compared to those with persistent restrictive filling
alone. Receiver operator characteristic analyses were
performed to determine the relationship between infarct
scar characteristics (IS and MVO) with adverse diastolic
remodeling and restrictive filling alone. Area under the
curve (AUC), sensitivity, and specificity were calculated.
For evaluation of the primary endpoint, Kaplan–Meier
survival analysis was performed, and differences were
assessed by the log-rank test. Univariate and multivariate
Cox proportional hazards regression models were
performed to determine significant predictors of the
composite adverse cardiovascular endpoints. The multi-
variate regression model comprised variables that were
shown to be significant at univariate analysis. The TIMI
risk score for STEMI (incorporating age, diabetes,
hypertension, heart rate, Killip class, weight, anterior
STEMI, and time to treatment) 16 was applied to the
multivariate model rather than individual demographic
characteristics, given the number of endpoints. Applica-
ble variables in our Cox proportional hazards regression
models were evaluated for multicollinearity. Given
collinearity between adverse diastolic remodeling and
 American Heart Journal
120 Nguyen et al October 2016

restrictive filling alone, the multivariate Cox regression Figure 1

models were derived with adverse diastolic remodeling
(model 1) and restrictive filling alone (model 2), substituted
into the respective model, with other variables kept
constant. All tests were 2-tailed, with a P b .05 considered
to be statistically significant. Statistical analyses were
performed with Statistical Package for Social Sciences
version 22.0 software (SPSS Inc, Chicago, IL).

Study approval
Study approval was obtained from the Human Research
Ethics Committee at Concord Hospital, Sydney Australia
(HREC/11/CRGH/224; approval CH62/6/2011–151).

Funding
TLN was supported by a co-funded National Health and
Medical Research Council and National Heart Foundation
postgraduate scholarship. The Cardiology department
has received educational grants from Biotronik, Medtro-
nic and St Jude Medical.
All other authors have no disclosures relevant to the
contents of this manuscript. The authors are solely
responsible for the design and conduct of this study, all
study analyses, the drafting and editing of the manuscript,
and its final contents.

Results
There were 409 consecutive STEMI patients screened
(May 2012 to June 2014). Of these, 265 patients were
enrolled (72 declined participation, 28 were non-English
speaking, 15 died prior to consent, 12 had chronic kidney
disease, 5 with previous cardiothoracic surgery, and 12
excluded for other causes). This cohort included patients
previously reported by our group. 7 A further 29 patients
were unable to undergo CMR, 12 later withdrew consent,
and 6 were lost to follow-up (Figure 1). Therefore a total
of 218 patients were considered in the final analysis. The
baseline characteristics of the study group are presented
in Table II.
Association of adverse diastolic remodeling with infarct
scar characteristics
The adverse diastolic remodeling group (n = 50) were
older and more likely males, with higher prevalence of
diabetes mellitus, hypertension, previous myocardial
infarction, and presentation with an anterior STEMI. By
receiver operator characteristic analysis, a follow-up IS of
9.7% could detect adverse diastolic remodeling with 78%
sensitivity and 82% specificity (AUC 0.86) (Figure 2A).
MVO was less predictive, compared with IS, for
identifying adverse diastolic remodeling (AUC 0.75).
When IS was grouped into tertiles, the upper IS tertile
group (both early and at follow-up) were significantly
associated with adverse diastolic remodeling (all
P b .001) (Figure 3).
Study flow diagram. *FU MACE includes 8 patients with early MACE
who have repeat event (after FU imaging study). Total MACE, n = 48.
CKD, chronic kidney disease; CMR, cardiac magnetic resonance
imaging; FU, follow-up; MACE, major adverse cardiovascular
event; STEMI, ST-segment elevation myocardial infarction; TTE,
transthoracic echocardiography.
Association of restrictive filling pattern and infarct scar
characteristics
Restrictive filling was more evident at follow-up (n =
33), versus early (n = 24) after STEMI. However, utiliza-
tion of the adverse diastolic remodeling criteria identified
a significantly higher percentage of patients (23%) than
presence of restrictive filling alone (11% early and 15% at
follow-up; both P b .001). We additionally sought to
determine the relationship between infarct scar charac-
teristics and restrictive filling. By receiver operator
characteristic analysis, follow-up IS (AUC 0.89), predicted
patients with persistent restrictive filling (Figure 2B), but
was less predictive for early restrictive filling (AUC 0.84).
A follow-up IS of 11.5% had 82% sensitivity and 86%
specificity in determining persistent restrictive filling.
Similar to adverse diastolic remodeling, MVO size was not
as predictive for identifying persistent restrictive filling
(AUC 0.76) compared with IS.
American Heart Journal
Volume 180
Nguyen et al 121

Table II. Baseline clinical and imaging characteristics of total study cohort, and patients stratified into no MACE and MACE groups
Total study group No MACE group MACE group

Parameter n = 218 n = 170 n = 48 P

Age, years⁎ 57.3 ± 10.6 56.4 ± 10.2 60.3 ± 11.6 .037

Gender, male 186 (85%) 149 (88%) 37 (77%) .068
BMI, m/kg2⁎ 27.4 ± 4.4 27.4 ± 4.6 27.5 ± 4.3 .864
eGRF, ml/min/1.73 m2⁎ 89 ± 26 90.9 ± 25.1 84.8 ± 29.7 .279
First presentation STEMI 199 (91%) 158 (93%) 41 (85%) .103
Diabetes mellitus 42 (19%) 25 (15%) 17 (35%) .001
Hypertension 100 (46%) 77 (45%) 23 (48%) .747
Anterior STEMI 117 (54%) 81 (48%) 36 (75%) .001
Killip class ≥2 33 (16%) 18 (11%) 15 (31%) b.001
TIMI risk score‡ 3 (2-5) 2 (1-4) 5 (3-7) b.001
Number of vessels with CAD†
Single vessel 141 (65%) 119 (70%) 22 (46%) b.001
Double vessel 59 (27%) 46 (28%) 13(24%)
Triple vessel 18 (8%) 5 (7%) 13 (14%)
Presentation
Primary PCI 174 (80%) 137 (81%) 37 (77%) .547
Thrombolysis + nonemergent PCI 29 (13%) 23 (14%) 6 (13%)
Thrombolysis + rescue PCI 15 (7%) 10 (6%) 5 (10%)
PCI success 213 (98%) 167 (98%) 46% (96%) .326
Medications on discharge
Aspirin 218 (100%) 170 (100%) 48(100%) 1.000
ADP antagonist 217 (99%) 169 (99%) 48(100%) .594
Beta blockers 205 (94%) 157 (92%) 48 (100%) .048
Statin 213 (98%) 165 (97%) 48 (100%) .229
ACEI/ARB 181 (83%) 140 (82%) 41 (85%) .618
Loop diuretic 16 (7%) 6 (4%) 10 (21%) b.001
Mineralocorticoid antagonist 13 (6%) 5 (3%) 8 (17%) b.001
Diastolic grade – early
0 – normal 38 (18%) 36 (21%) 2 (4%) b.001
1 – impaired relaxation 79 (36%) 64 (38%) 15 (31%)
2 – pseudonormal 77 (35%) 60 (35%) 17 (36%)
3 – restrictive filling 24 (11%) 10 (6%) 14 (29%)
Diastolic grade – follow-up
0 – normal 55 (25%) 48 (28%) 7 (14%) b.001
1 – impaired relaxation 77 (35%) 69 (41%) 8 (17%)
2 – pseudonormal 53 (25%) 44 (26%) 9 (19%)
3 – restrictive filling 33 (15%) 9 (5%) 24 (50%)
Adverse diastolic remodeling 50 (23%) 19 (11%) 31 (65%) b.001
CMR characteristics
Presence of MVO 95 (44%) 64 (38%) 31 (66%) .001
MVO size, %‡ 0.00 (0.00-0.45) 0.00 (0.00-0.36) 0.13 (0.00-1.15) .001
Early infarct scar size, %‡ 8.9 (5.3-13.9) 8.2 (5.2-12.6) 13.7 (8.1-22.7) b.001
Early LV ejection fraction, %‡ 46.4 (38.8-52.8) 47.4 (41.8-53.1) 37.6 (30.6-49.1) b.001
FU infarct scar size, %‡ 7.0 (4.1-11.5) 6.6 (3.9-9.2) 12.6 (6.4-21.7) b.001
FU LV ejection fraction, %‡ 52.4 (42.0-58.8) 53.8 (47.0-59.6) 40.0 (28.1-51.8) b.001

ACEI, angiotensin-converting enzyme inhibitor; ADP, adenosine diphosphate; ARB, angiotensin receptor blocker; BMI, body mass index; CAD, coronary artery disease; CMR, cardiac
magnetic resonance imaging; eGFR, estimated glomerular filtration rate; LV, left ventricular; MVO, microvascular obstruction; PCI, percutaneous coronary intervention; STEMI,
ST-segment elevation myocardial infarction.
⁎ mean ± SD.
† coronary artery stenosis ≥70%.
‡ median (interquartile range);

Determinants of clinical outcomes
Over a mean follow up of 710 ± 79 days there were 61
primary events (including 5 deaths, 7 episodes of spontane-
ous ventricular arrhythmias, 23 reinfarctions, 21 heart failure
exacerbations, and 5 cerebrovascular accidents) in 48
patients. In an unadjusted Kaplan Meier survival analysis,
patients with adverse diastolic remodeling or persistent
restrictive filling alone had similarly significant outcomes for
MACE (both P b .001) (Figure 4). The patient cohort,
stratified into those without and with MACE groups, are
presented in Table II. The groups had similar usage of
antiplatelet agents, statins, and angiotensin converting
enzyme receptor inhibitor/angiotensin II receptor blockers.
There were a higher percentage of patients on beta blockers,
 American Heart Journal
122 Nguyen et al October 2016

Figure 2

Receiver operator characteristic analyses evaluating the relationship of infarct scar size and MVO size with A) adverse diastolic remodeling, and B)
persistent restrictive filling. AUC, area under the curve; CI, confidence interval; FU, follow-up; MVO, microvascular obstruction.

Figure 3

Relationship of infarct scar size tertile groups (at follow-up) and adverse diastolic remodeling. *χ 2 analysis, infarct scar size tertile 1 vs tertile 3;
infarct scar size tertile 2 vs tertile 3; and infarct scar size tertiles 1 and 2 vs tertile 3; all P b .001.
American Heart Journal
Volume 180
Figure 4
Kaplan Meier survival analyses for adverse diastolic remodeling, and persistent restrictive filling with major adverse cardiovascular events.
loop diuretics and mineralocorticoid receptor antagonists in
the group with MACE. Similarly when the groups were
stratified into those with adverse diastolic remodeling, the
use of beta blockers (100% vs 92%, P = .043), loop diuretic
(22% vs 3%, P b .001) and mineralocorticoid antagonists
(18% vs 2%, P b .001) were higher when compared to
patients without adverse diastolic remodeling.
By Cox proportional hazards regression analysis,
persistent restrictive filling and adverse diastolic remod-
eling were univariate predictors for MACE (Table III).
MVO size, IS and LV ejection fraction at follow-up, and
the change in LV ejection fraction, were also univariate
predictors of MACE. In addition, the TIMI risk score was a
univariate predictor of adverse cardiovascular outcomes.
After exclusion of variables with significant multicolli-
nearity (Pearson correlation r ≥ 0.8), the multivariate
Cox proportional hazards regression model was per-
formed using significant univariate predictors. Given that
there was collinearity between adverse diastolic remod-
eling and persistent restrictive filling (r = 0.8), 2 separate
models were tested by substituting adverse diastolic
remodeling or persistent restrictive filling into the final
multivariate Cox regression model, whilst other covari-
ates remained constant. The variables entered in the
multivariate Cox regression model included: follow-up IS,
MVO size, follow-up LV ejection fraction, TIMI risk score,
and adverse diastolic remodeling (model 1) or persistent
restrictive filling (model 2). Adverse diastolic remodeling
(HR 3.79, P b .001), and TIMI risk score (HR 1.24, P =
Nguyen et al 123
.01), remained the only independent predictors for the
primary outcome of MACE (Table IV). In the second
multivariate model when persistent restrictive filling was
substituted for adverse diastolic remodeling, there was a
significant association for MACE (HR 2.61 P = .019). Both
follow-up LV ejection fraction and IS were non-significant
predictors in the model. Adverse diastolic remodeling
remained highly significant (P b .001) when early param-
eters (IS, MVO and LVEF) or LV systolic remodeling were
substituted into the multivariate models (Supplemental
Tables I and II).
Discussion
This study, whilst reaffirming the value of restrictive
filling as a prognostic predictor for MACE, has demon-
strated the additional prognostic value of adverse
diastolic remodeling by serial diastolic assessment after
STEMI (Figure 5). IS, and to a lesser extent MVO size, are
associated with adverse diastolic remodeling. As the LV
undergoes remodeling post STEMI, serial diastolic evalu-
ation is better able to determine poor cardiovascular
outcomes, than a single time point analysis. Both adverse
diastolic remodeling and persistent restrictive filling were
significant predictors for MACE. However, adverse
diastolic remodeling had a stronger association with
MACE and identified a larger group of patients compared
to restrictive filling alone when adjusted for TIMI risk
score, IS, MVO size and LV ejection fraction.
 American Heart Journal
124 Nguyen et al October 2016

Table III. Univariate Cox proportional hazards regression Table IV. Multivariate Cox proportional hazards regression
analysis of predictors for major adverse cardiovascular events analyses of independent predictors for major adverse
cardiovascular events
χ2 HR P
χ2 HR P
MVO size, % 5.89 1.16 (1.03-1.30) .016
Early infarct scar size, % 30.82 1.10 (1.07-1.14) b.001 Model 1: adverse diastolic remodeling⁎
Early CMR LV ejection fraction, % 32.75 0.91 (0.87-0.94) b.001 Adverse diastolic remodeling 12.49 3.79 (1.81-7.93) b.001
FU infarct scar size, % 50.27 1.14 (1.10-1.18) b.001 TIMI risk score 6.70 1.24 (1.05-1.45) .010
FU CMR LV ejection fraction, % 49.33 0.91 (0.89-0.94) b.001 FU infarct scar size, % 2.97 1.04 (0.99-1.10) .085
Change in LV ejection fraction, %⁎ 13.79 0.89 (0.83-0.95) b.001 Model 2: persistent restrictive filling⁎
Adverse diastolic remodeling 52.30 9.01 (4.97-16.35) b.001 Persistent restrictive filling 5.50 2.61 (1.71-5.83) .019
Persistent restrictive filling 56.08 8.91 (5.03-15.80) b.001 TIMI risk score 6.67 1.24 (1.05-1.46) .010
Age, years 5.68 1.04 (1.01-1.07) .017 FU LV ejection fraction, % 3.72 0.97 (0.93-1.00) .054
Diabetes mellitus 9.68 2.56 (1.42-4.63) .002
⁎ Variables in multivariate Cox model: TIMI risk score, FU LV ejection fraction,
Anterior STEMI 10.26 2.91 (1.51-5.59) .001
microvascular obstruction size, FU infract scar size, and adverse diastolic remodeling
Previous myocardial infarction 4.80 2.49 (1.10-5.63) .029 (model 1) or persistent restrictive filling (model 2). FU, follow-up; HR, hazard ratio; LV,
Hypertension 0.23 1.15 (0.65-2.03) .63 left ventricular.
Killip class 28.79 1.99 (1.55-2.55) b.001
TIMI risk score 49.31 1.54 (1.37-1.74) b.001 with thrombolysis, and only 48% underwent an angiogram
⁎ Change from early to follow up scans. CMR, cardiac magnetic resonance imaging; within 48 hours. Similar to our report, cumulative 1 year
FU, follow-up; HR, hazard ration; LV, left ventricular; MVO, microvascular obstruction; outcomes of heart failure and cardiac death, was significantly
STEMI, ST-segment elevation myocardial infarction.
worse in the group with persistence or deterioration in
abnormal filling (P b .0001). Moller et al also reported that
Prognostic significance of diastolic function parameters patients post acute myocardial infarction with pseudonor-
Restrictive filling is a well described prognostic marker mal or restrictive filling pattern were more likely to suffer
after both STEMI and non-STEMI. 4,17-20 In the largest cardiac death. 19 However, these studies were performed in
study by the MeRGE investigators, a metanalysis of 12 the pre-PCI era, with most recruited patients being treated
prospective studies (n = 3396 patients), demonstrated with thrombolysis. Our study pertains to more contempo-
the prognostic value of restrictive filling. 4 The persis- rary practice with respect to STEMI treatment, with 80%
tence of restrictive filling on serial imaging, by Temporelli treated by primary PCI, but additionally included evaluation
et al, 21 identified larger LV end systolic volumes and of scar size by CMR that ascertains the extent of LV
predicted long-term mortality at 4 years. However, the myocardial damage more accurately.
majority of STEMI patients were treated with thrombo- Patients with previous myocardial infarction (9%) were
lytic therapy. In addition, the reversibility of baseline included in our study cohort, as we examined serial change
restrictive filling to a non-restrictive pattern during early in diastolic function, regardless of baseline diastolic dysfunc-
follow-up suggested a more favorable outcome. Our tion. Increased LV filling pressure and diastolic dysfunction
study confirms the prognostic value of persistent reflect increased risk in those with cardiovascular comor-
restrictive filling (within the first 2 months post STEMI) bidities, including previous myocardial infarction, despite
for MACE, even in STEMI patients treated primarily by preserved or mildly impaired systolic function. 5 Although
contemporary PCI based revascularization (Figure 4). restrictive filling remains a prognostic marker, in the
However, restrictive filling alone only occurred in a small contemporary era of primary PCI treatment for STEMI, we
number of study patients (11% early and 15% at have demonstrated that adverse diastolic remodeling
follow-up), and identified a significantly smaller number identifies a larger group of ‘at risk patients’ emphasizing
of patients at risk of adverse events, compared to adverse the value of examining serial change in diastolic dysfunction,
diastolic remodeling (23%; both P b .001). and additionally demonstrating a stronger association with
There are only a limited number of studies that have MACE than persistent restrictive filling pattern.
performed serial evaluation of diastolic function post STEMI,
particularly in patients treated by PCI. Poulsen et al reported Diastolic function and infarct scar characteristics
serial changes in diastolic grade up to 1 year, but 72% of the The pathophysiology of diastolic dysfunction after
patients in this study were treated with thrombolytic myocardial infarction involves several factors including
therapy. 22 They reported both increased heart failure events increased LV filling pressure, impaired active LV chamber
as well as death in patients with pseudonormal or restrictive relaxation and stiffness resulting from myocardial dam-
filling patterns but did not specifically examine changes in age. 5 Myocardial ischemia, myocyte necrosis and micro-
diastolic function grade. Moller et al evaluated changes in LV vascular injury with resulting regional wall motion
diastolic filling patterns (between baseline and echocardio- abnormalities can also adversely influence the rate of
gram performed at 1 month) on 1 year prognosis, in active LV relaxation and remodeling. 24 Furthermore,
103 patients with first presentation acute myocardial interstitial edema and infarct scar formation contribute to
infarction. 23 Again, 54% of recruited subjects were treated the degree of LV stiffness.
American Heart Journal
Volume 180
Nguyen et al 125

Figure 5

Diagrammatic representation of measurements for diastolic function (pulse wave Doppler, and tissue Doppler imaging) and late gadolinium
enhancement cardiac magnetic resonance imaging infarct characteristics (infarct scar size = red arrow, and microvascular obstruction = blue
arrow) in a patient with an anterior STEMI. Diastolic function grade measurements show a restrictive filling pattern.

IS is a significant prognostic marker, beyond LV
ejection fraction and LV end systolic volumes. 25,26 MVO
too has been a determinant of LV remodeling and
prognosis. 1,27 Stratifying patients into those who had
adverse diastolic remodeling, provided a good association
with IS and to a lesser extent MVO size. Furthermore,
adverse diastolic remodeling occurred particularly in
patients with the larger IS (upper tertile) both early and at
follow-up. There have been only a small number of
studies that have evaluated the relationship of diastolic
function grade and IS. Barbieri et al demonstrated, in 52
subjects, a weak but significant correlation of diastolic
function grade to ‘healed’ infarct size, measured by late
gadolinium enhancement CMR, within 2 months after
STEMI. 28 However, similar to our findings, among
patients with advanced ischemic cardiomyopathy, dia-
stolic dysfunction was associated with increasing myo-
cardial scar burden. 29
Adverse diastolic remodeling is also linked to LV systolic
function and extent of myocardial damage after STEMI. As
shown in Table II, there is significantly larger IS and MVO, in
those with adverse diastolic remodeling. Furthermore the
recovery of IS and LV ejection fraction from baseline to
follow up is reduced in those with adverse diastolic
remodeling. Although the extent of improvement in LV
ejection fraction within the first 2 months was shown to be
negatively associated with MACE, the prognostic value of
adverse diastolic remodeling was independent of the change
in LV ejection fraction.
Both the presence of MVO and extent, although
significantly higher in those with adverse diastolic
remodeling, was less predictive for diastolic dysfunction
compared to IS. This can be explained by the several
mechanisms involved in the pathophysiology of MVO
including ischemia and reperfusion injury, distal coronary
embolization, and individual patient susceptibility. 30

126 Nguyen et al
Although MVO was associated with systolic segmental
dysfunction, in a canine model, MVO was also related to
persistent regional diastolic impairment after STEMI. 31
This was also demonstrated in our results showing an
association of MVO with adverse diastolic remodeling
and persistent restrictive filling.
Study limitations
This was an observational study; however, consecutive
patients were prospectively recruited and evaluated. We
included only patients who had serial TTE and CMR scans,
and consented to long-term follow-up. Thus, those with
qualifying imaging were somewhat lower risk so the
reported influence of adverse diastolic remodeling on
MACE may indeed be under-estimated. Patients with
moderate to severe chronic kidney disease, an independent
marker of adverse diastolic function, 32 were not included
in this study due to the adverse risk of intravenous
gadolinium administration for CMR. Owing to improved
revascularization strategies, there has been greater patient
survival post STEMI in recent years. 33 A composite
endpoint of mortality and other major cardiovascular
events was studied, given the low mortality (5 deaths,
2.2%) and the relatively short duration of follow up. While
rates of evidence based medical therapy at discharge were
satisfactory (with the possible exception of aldosterone
antagonists), the durations of specific medications after
discharge were not routinely recorded. Given this, and the
limited number of patients with adverse diastolic remod-
eling (n = 50), we therefore could not determine if
duration of medication would have had a significant impact
on clinical outcomes.
Conclusions
Larger IS results in adverse diastolic remodeling,
demonstrating the link between extent of scar and
myocardial damage and serially evaluated diastolic
function. Following contemporary management of
STEMI, identification of adverse diastolic remodeling is
a powerful prognostic marker and identifies a larger
group of ‘at-risk’ patients than persistent restrictive filling
alone. Hence, this emphasizes the importance of serial
monitoring of diastolic function after STEMI.
Perspectives
Competency in medical knowledge. Diastolic
dysfunction after STEMI is a dynamic process. Adverse
diastolic remodeling reflects worsening LV structural
remodeling, reduced LV function, larger IS, and poor
myocardial viability. This consequently has adverse
prognostic implications.
Translational outlook. Further studies are required
to assess the impact of long-term medical therapy on
adverse diastolic remodeling, and consequent changes in
clinical outcomes after STEMI.
American Heart Journal
October 2016
Appendix. Supplementary data
Supplementary data to this article can be found online
at http://dx.doi.org/10.1016/j.ahj.2016.05.020.
References
1. van Kranenburg M, Magro M, Thiele H, et al. prognostic value of
microvascular obstruction and infarct size, as measured by CMR in
STEMI patients. J Am Coll Cardiol Img 2014;7:930-9.
2. El Aidi H, Adams A, Moons KGM, et al. Cardiac magnetic resonance
imaging findings and the risk of cardiovascular events in patients with
recent myocardial infarction or suspected or known coronary artery
disease. J Am Coll Cardiol 2014;63:1031-45.
3. Whalley GA, Gamble GD, Dini FL, et al. Individual patient
meta-analyses of restrictive diastolic filling pattern and mortality in
patients post acute myocardial infarction and in patients with chronic
heart failure. Int J Cardiol 2007;122:207-15.
4. Meta-Analysis Research Group in Echocardiography (MeRGE) AMI
CollaboratorsMøller JE, Whalley GA, et al. Independent prognostic
importance of a restrictive left ventricular filling pattern after
myocardial infarction: an individual patient meta-analysis:
Meta-Analysis Research Group in Echocardiography acute myocar-
dial infarction. Circulation 2008;117:2591-8.
5. Moller JE, Pellikka PA, Hillis GS, et al. Prognostic importance of
diastolic function and filling pressure in patients with acute myocardial
infarction. Circulation 2006;114:438-44.
6. Husic M, Norager B, Egstrup K, et al. Serial changes in regional
diastolic left ventricular function after a first acute myocardial
infarction. J Am Soc Echocardiogr 2005;18:1173-80.
7. Nguyen TL, Phan JA, Hee L, et al. High sensitivity Troponin T predicts
infarct scar characteristics and adverse left ventricular function by
cardiac magnetic resonance imaging early after reperfused acute
myocardial infarction. Am Heart J 2015;170:715-25.
8. Nagueh SF, Appleton CP, Gillebert TC, et al. Recommendations for
the evaluation of left ventricular diastolic function by echocardiogra-
phy. Eur J Echocardiogr 2009;10:165-93.
9. Schulz-Menger J, Bluemke DA, Bremerich J, et al. Standardized
image interpretation and post processing in cardiovascular magnetic
resonance: Society for Cardiovascular Magnetic Resonance (SCMR)
Board of Trustees Task Force on Standardized Post Processing. J
Cardiovasc Magn Reson 2013;15.
10. Gao P, Yee R, Gula L, et al. Prediction of arrhythmic events in ischemic
and dilated cardiomyopathy patients referred for implantable cardiac
defibrillator: evaluation of multiple scar quantification measures for
late gadolinium enhancement magnetic resonance imaging. Circ
Cardiovasc Imaging 2012;5:448-56.
11. Eitel I, de Waha S, Wöhrle J, et al. Comprehensive prognosis
assessment by CMR imaging after ST-segment elevation myocardial
infarction. J Am Coll Cardiol 2014;64:1217-26.
12. Mayr A, Klug G, Schocke M, et al. Late microvascular obstruction
after acute myocardial infarction: Relation with cardiac and
inflammatory markers. Int J Cardiol 2012;157:391-6.
13. Thygesen K, Alpert JS, Jaffe AS, et al. Third universal definition of
myocardial infarction. Circulation 2012;126:2020-35.
14. Zaman S, Sivagangabalan G, Narayan A, et al. Outcomes of early
risk stratification and targeted implantable cardioverter-defibrillator
implantation after ST-elevation myocardial infarction treated with
primary percutaneous coronary intervention. Circulation 2009;120:
194-200.
American Heart Journal
Volume 180
15. Richards DA, Byth K, Ross DL, et al. What is the best predictor of
spontaneous ventricular tachycardia and sudden death after myo-
cardial infarction? Circulation 1991;83:756-63.
16. Morrow DA, Antman EM, Charlesworth A, et al. TIMI Risk Score for
ST-Elevation Myocardial Infarction: A Convenient, Bedside, Clinical
Score for Risk Assessment at Presentation: An Intravenous nPA for
Treatment of Infarcting Myocardium Early II Trial Substudy. Circula-
tion 2000;102:2031-7.
17. Nagueh SF, Middleton KJ, Kopelen HA, et al. Doppler tissue imaging:
a noninvasive technique for evaluation of left ventricular relaxation
and estimation of filling pressures. J Am Coll Cardiol 1997;30:
1527-33.
18. Nijland F, Kamp O, Karreman AJ, et al. Prognostic implications of
restrictive left ventricular filling in acute myocardial infarction: a serial
Doppler echocardiographic study. J Am Coll Cardiol 1997;30:
1618-24.
19. Møller JE, Søndergaard E, Poulsen SH, et al. Pseudonormal and
restrictive filling patterns predict left ventricular dilation and cardiac
death after a first myocardial infarction: a serial color M-mode
Doppler echocardiographic study. J Am Coll Cardiol 2000;36:
1841-6.
20. Prasad SB, See V, Brown P, et al. Impact of duration of ischemia on left
ventricular diastolic properties following reperfusion for acute
myocardial infarction. Am J Cardiol 2011;108:348-54.
21. Temporelli PL, Giannuzzi P, Nicolosi GL, et al. Doppler-derived mitral
deceleration time as a strong prognostic marker of left ventricular
remodeling and survival after acute myocardial infarction: results of
the GISSI-3 echo substudy. J Am Coll Cardiol 2004;43:1646-53.
22. Poulsen SH, Jensen SE, Egstrup K. Longitudinal changes and
prognostic implications of left ventricular diastolic function in first
acute myocardial infarction. Am Heart J 1999;137:910-8.
23. Møller JE, Poulsen SH, Søndergaard E, et al. Impact of early changes
in left ventricular filling pattern on long-term outcome after acute
myocardial infarction. Int J Cardiol 2003;89:207-15.
Nguyen et al 127
24. Sutton MG, Sharpe N. Left ventricular remodeling after myocardial
infarction: pathophysiology and therapy. Circulation 2000;101:
2981-8.
25. Roes SD, Kelle S, Kaandorp TAM, et al. Comparison of myocardial infarct
size assessed with contrast-enhanced magnetic resonance imaging and
left ventricular function and volumes to predict mortality in patients with
healed myocardial infarction. Am J Cardiol 2007;100:930-6.
26. Kelle S, Roes SD, Klein C, et al. Prognostic value of myocardial infarct
size and contractile reserve using magnetic resonance imaging. J Am
Coll Cardiol 2009;54:1770-7.
27. Hamirani YS, Wong A, Kramer CM, et al. Effect of microvascular
obstruction and intramyocardial hemorrhage by CMR on LV
remodeling and outcomes after myocardial infarction. J Am Coll
Cardiol Img 2014;7:940-52.
28. Barbieri A, Bursi F, Politi L, et al. Echocardiographic diastolic
dysfunction and magnetic resonance infarct size in healed myocardial
infarction treated with primary angioplasty. Echocardiography
2008;25:575-83.
29. Cavalcante JL, Marwick TH, Hachamovitch R, et al. Is there a role for
diastolic function assessment in era of delayed enhancement cardiac
magnetic resonance imaging? Am Heart J 2014;168:220-8. [e1].
30. Niccoli G, Burzotta F, Galiuto L, et al. Myocardial No-Reflow in
Humans. J Am Coll Cardiol 2009;54:281-92.
31. Azevedo C, Amado L, Kraitchman D, et al. Persistent diastolic
dysfunction despite complete systolic functional recovery after
reperfused acute myocardial infarction demonstrated by tagged
magnetic resonance imaging. Eur Heart J 2004;25:1419-27.
32. Farshid A, Pathak R, Shadbolt B, et al. Diastolic function is a strong
predictor of mortality in patients with chronic kidney disease. BMC
Nephrol 2013;14:280.
33. Goldberg RJ, Spencer FA, Gore JM, et al. Thirty-Year Trends (1975 to
2005) in the magnitude of, management of, and hospital death rates
associated with cardiogenic shock in patients with acute myocardial
infarction: a population-based perspective. Circulation 2009;119:1211-9.