DOI 10.1378/chest.09-1996
2010;138;198-207Chest
Dellsperger
Anand Chockalingam, Ankit Mehra, Smrita Dorairajan and Kevin C.
Critically Ill
Acute Left Ventricular Dysfunction in the
http://chestjournal.chestpubs.org/content/138/1/198.full.html
services can be found online on the World Wide Web at:
The online version of this article, along with updated information and
ISSN:0012-3692
)http://chestjournal.chestpubs.org/site/misc/reprints.xhtml(
written permission of the copyright holder.
this article or PDF may be reproduced or distributed without the prior
Dundee Road, Northbrook, IL 60062. All rights reserved. No part of
Copyright2010by the American College of Chest Physicians, 3300
Physicians. It has been published monthly since 1935.
is the official journal of the American College of ChestChest
© 2010 American College of Chest Physicians
by guest on September 8, 2011chestjournal.chestpubs.orgDownloaded from
Postgraduate Education Corner
CONTEMPORARY REVIEWS IN CRITICAL CARE MEDICINE
CHEST
198 Postgraduate Education Corner
Acute left ventricular (LV) dysfunction occurs in about one-third of critically ill hospitalized
patients. 1 - 4 The increasing incidence of LV dysfunc-
tion in ICUs is likely related to both changing patient
characteristics (advancing age, increased comorbidi-
ties) and practice patterns (widespread troponin, cre-
atine kinase-MB, and brain natriuretic peptide [BNP]
testing, as well as more frequent performance of
bedside echocardiography). 3 , 5 - 8 A determination as
to whether the LV dysfunction is the cause, effect, or
a coincidental fi nding has to be made and revisited
periodically. Acute medical or surgical plans, ongoing
management targets, outcome expectations, and
prognosis must be reconciled. Recognizing that all
the individual causes and complexities cannot be cap-
tured here, we will summarize the most important
causes of LV dysfunction in the critically ill ( Table 1 )
and present a unifi ed management approach from
the cardiac standpoint.
Diagnosis of LV Dysfunction
Angina, dyspnea, pulmonary crackles, murmurs,
tachyarrhythmias, biomarker elevations, or ischemic
ECG changes suggest cardiac pathology in hospital-
ized patients. Because of variability in patient charac-
teristics and study design, predictive values of each
cardiac test remain unclear. Several studies suggest
routine troponin screening in ICUs may be sensitive
in detecting early cardiac involvement among the
critically ill. 9 , 10 Supporting this, 15% to 30% of critically
ill noncardiac patients develop troponin elevations
and this corresponds with poorer outcomes. Overall
Acute Left Ventricular Dysfunction
in the Critically Ill
Anand Chockalingam , MD ; Ankit Mehra , MD ; Smrita Dorairajan , MD ;
and Kevin C. Dellsperger , MD, PhD
Manuscript received August 22 , 2009 ; revision accepted December
7 , 2009 .
Affi liations: From the Division of Cardiovascular Medicine,
Department of Internal Medicine (Drs Chockalingam, Mehra,
Dorairajan, and Dellsperger), University of Missouri School of
Medicine; and the Cardiology Section (Dr Chockalingam), Harry
S. Truman VA Medical Center, Columbia, MO.
Funding�Support: This study was funded by Veterans Adminis-
tration Research award [VISN 15] to Dr Chockalingam.
Correspondence to: Anand Chockalingam, MD, Division of
Cardiology, University of Missouri–Columbia, 5 Hospital Dr,
CE306, Columbia, MO 65212; e-mail: chockalingama@health.
missouri.edu
© 2010 American College of Chest Physicians. Reproduction
of this article is prohibited without written permission from the
American College of Chest Physicians ( www . chestpubs . org �
site � misc � reprints . xhtml ).
DOI: 10.1378/chest.09-1996
Acute left ventricular (LV) dysfunction is common in the critical care setting and more frequently
affects the elderly and patients with comorbidities. Because of increased mortality and the poten-
tial for signifi cant improvement with early revascularization, the practitioner must fi rst consider
acute coronary syndrome. However, variants of stress (takotsubo) cardiomyopathy may be more
prevalent in ICU settings than previously recognized. Early diagnosis is important to direct treat-
ment of complications of stress cardiomyopathy, such as dynamic LV outfl ow tract obstruction,
heart failure, and arrhythmias. Global LV dysfunction occurs in the critically ill because of the
cardio-depressant effect of infl ammatory mediators and endotoxins in septic shock as well as
direct cate cholamine toxicity. Tachycardia, hypertension, and severe metabolic abnormalities can
independently cause global LV dysfunction, which typically improves with addressing the precipi-
tating factor. Routine troponin testing may help early detection of cardiac injury and biomarkers
could have prognostic value independent of prior cardiac disease. Echocardiography is ideally suited
to quantify LV dysfunction and determine its most likely cause. LV dysfunction suggests a worse
prognosis, but with appropriate therapy outcomes can be optimized. CHEST 2010; 138( 1 ): 198 – 207
Abbreviations : ACS 5 acute cardiac syndrome ; BNP 5 brain natriuretic peptide ; LV 5 left ventricular ; NT-proBNP 5 N-terminal-
pro-B-type natriuretic peptide ; RWMA 5 regional wall motion abnormalities
© 2010 American College of Chest Physicians
by guest on September 8, 2011chestjournal.chestpubs.orgDownloaded from
CHEST / 138 / 1 / JULY, 2010 199www.chestpubs.org
mortality in one study was 27% (58 of 217 patients),
but patients with troponin elevation had a much
higher mortality (51%) compared with those without
(16%). 10 Although routine troponin testing may help
identify LV dysfunction early, there is currently no
evidence that this improves outcomes in critical care
patients.
Signifi cant elevation in plasma levels of N-terminal-
pro-B-type natriuretic peptide concentrations
(NT-proBNP) and BNP are typically diagnostic of car-
diac pathology as cause for dyspnea and heart fail ure.
In critically ill patients with shock, however, BNP tends
to be elevated and is thus not reliable for diagnosing
heart failure. 11 , 12 Because BNP is higher in sepsis non-
survivors (943 pg�mL vs 378 pg�mL in survivors), some
believe it may play a prognostic role. 13
Transthoracic echocardiography, being portable,
noninvasive, and easily repeatable, is ideal to evaluate
LV dysfunction in critical care settings. 3 , 5 In addition,
right ventricular function, pulmonary pressures, valve
disease, and pericardial pathology, along with hemo-
dynamic parameters, such as central volume status
and cardiac output, can be quantifi ed and serially
monitored. 6 About 8% to 20% of critically ill patients
manifest LV dysfunction, although one serial echo-
Table 1 —Causes of Acute Left Ventricular Dysfunction
in the Critical Care Setting
Myocardial infarction
Typical acute coronary syndromes
Signifi cant myocardial involvement
Mechanical complications, such as ventricular septal rupture
and papillary muscle rupture
Paradoxical venous thromboembolism
Coronary emboli—left atrial myxoma, LV or atrial thrombus
Coronary thrombosis—antiphospholipid antibody syndrome,
disseminated intravascular coagulation, thrombotic
thrombocytopenic purpura
Aortic dissection—with right coronary occlusion
Stress cardiomyopathy
Apical ballooning 6 LV outfl ow obstruction
Basal cardiomyopathy (apex-sparing)
Focal cardiomyopathy (noncoronary distribution)
Global hypokinesis
Tachyarrhythmias
Hypertensive emergency
Sepsis
Metabolic and multiorgan insults
Post cardiac arrest and resuscitation
Myocardial injury with minor troponin elevations (including
supply-demand mismatch)
Myopericarditis—viral, autoimmune, giant cell
Trauma—chest contusion, prolonged resuscitation, bleeding
Congestive heart failure—decompensated, with anemia, shunts
Pulmonary embolism—with right ventricle strain
Sepsis—hypotension, catecholamine drips
Extracardiac stressors—hypertensive crisis, thyrotoxicosis, cocaine,
hypothermia, drowning
Prolonged surgery—hypotension, blood loss
LV 5 left ventricular.
cardiographic study suggests a higher incidence, up
to 30%. 2 Importantly, presence, extent, and location
of regional wall motion abnormalities as well as LV
dimensions and shape help determine the most likely
cause for LV dysfunction and guide further manage-
ment in most instances.
Cardiac catheterization provides defi nitive assess-
ment of coronary disease. CT scan angiography and
cardiac MRI may be valuable in specifi c situations
but are limited in ICU settings. 14 Table 2 summarizes
key diagnostic advantages and limitations with var-
ious cardiac tests in the critical care setting.
Acute Coronary Syndromes
Diagnosis
Plaque rupture resulting in total occlusion of a
major coronary artery results in chest pain with ECG
evidence of ST elevation. This is associated with a siz-
able territory of myocardium in jeopardy. This war-
rants emergent coronary angiography and revascular-
ization. Chest pain with diaphoresis and dyspnea may
be reported but classic symptoms may be masked in
many ICU patients because of sedation or altered
mental status. 3 , 5 Hemodynamic changes, such as hypo-
tension, low cardiac output, reduced mixed venous
saturation, and increasing pulmonary wedge pressures
may trigger performing a 12-lead ECG in the sedated
or unconscious patient. Increasing ventricular ectopy,
ST segment changes, or new bundle branch blocks on
telemetry may also be the initial evidence for cardiac
ischemia. 1 , 15 When clinical and ECG fi ndings are
equivocal, bedside echocardiography with careful
evaluation for regional wall motion abnormalities
(RWMA) conforming to typical coronary distributions
may help to confi rm acute coronary syndrome (ACS). 3 , 6 , 9
Mechanical complications of ACS, such as mitral
regurgitation or ventricular septal defects, may also
be detected with echocardiography. 16
ACS Without Critical Coronary Artery Disease
Medical, surgical, or trauma intensive care may
result in substantial physiologic and mental stress.
This may cause alterations in hemodynamics, coagula-
bility, and metabolic parameters. Patients with preex-
isting signifi cant atherosclerotic stenosis may not be
able to increase blood supply commensurate with the
increasing demands resulting in supply-demand mis-
match. This usually manifests as ACS with ischemic
ECG changes and troponin elevations. Although new
LV dysfunction is not typical with supply-demand
mismatch, it is included in this review because of the
relatively higher incidence of supply-demand mis-
match in critical care settings. 17 - 19 This may account
© 2010 American College of Chest Physicians
by guest on September 8, 2011chestjournal.chestpubs.orgDownloaded from
200 Postgraduate Education Corner
for troponin elevations seen in about one-fourth of
critically ill noncardiac patients ( Table 1 ). 9 , 10
In instances without angiographic culprit lesions,
in situ coronary arterial thrombosis due to hyperco-
agulable conditions, such as thrombocytosis, dissem-
inated intravascular coagulation, thrombocytopenic
purpura, and antiphospholipid antibody syndrome,
need to be considered. 20 , 21 Rarely embolic coronary
occlusion may be due to left-side heart (atrial or ventric-
ular) mural thrombi, endocarditis, prosthetic valve
thrombi, or cardiac myxoma. 22 Paradoxical emboli
and thrombi from intracardiac catheters or guide-
wires have to be considered in patients with a patent
foramen ovale. 23 Therapy is mainly supportive and
aimed at preventing recurrence by addressing pre-
cipitation factors. This might involve surgery for car-
diac tumors, percutaneous device closure of patent
foramen ovale, or anticoagulation for hypercoagu-
lable states. Cardiac status may potentially improve
with antiplatelet (aspirin, clopidogrel, glycoprotein
IIb�IIIa antagonists) and anticoagulant therapy (hep-
arin, warfarin). However there are no studies in crit-
ical care settings for these therapies because of wide
variability in cause for ACS and comorbidities. With
the higher bleeding risk in this population, these
agents must be used on a case-by-case basis.
Stress Cardiomyopathy
Defi nition and Epidemiology
Originally described in Japan as takotsubo cardio-
myopathy, stress cardiomyopathy by defi nition implies
completely reversible acute LV dysfunction. 24 - 28 ICU
admission because of medical illness, surgical proce-
dure, or traumatic injuries could typically be suffi cient
stress to cause stress cardiomyopathy. 29 By perform-
ing serial echocardiography in consecutive ICU
patients, 28% (26 of 92 patients) had stress cardiomy-
opathy in one series reported from South Korea. 2
This is much higher than we typically encounter and
a more recent larger series in which echocardiograms
were obtained routinely in the fi rst 24 h of ICU
admission detected LV systolic dysfunction in 132
of 704 (18%) patients. 3
Pathogenesis of Stress Cardiomyopathy
Catecholamine excess in circulation has been iden-
tifi ed and possibly mediates the acute cardiac dys-
function in stress cardiomyopathy. 30 This condition
typically affects women (in . 80% of most series) in
the 62- to 75-year-old age range. 31 , 32 Severe emotional
stress (approximately 27%) or physical illness (approx-
imately 38%), such as sepsis, head trauma, and cere-
brovascular accident, may precipitate stress cardio-
myopathy in about two-thirds of the instances. 32 In
about 60% to 80% of those with this condition the
mid and distal segments of the LV are akinetic with a
hypercontractile base giving the appearance of “apical
ballooning.” Stress cardiomyopathy can also affect
the base with apical sparing in another 10%. The
remainder manifests nonspecifi c regional wall motion
abnormalities or global hypokinesia not conforming
to any particular coronary territory. 28 , 33 This global
hypokinesia variant of stress cardiomyopathy is unlikely
Table 2 —Summary of Various Cardiac Tests Available Highlighting Their Key Characteristics in Critical
Care Settings
Test Key Diagnosis Advantages Limitations
ECG ST elevation MI Widely available Stress cardiomyopathy may mimic MI
Non-ST elevation MI Inexpensive
Arrhythmias
Troponin Any myocardial injury High sensitivity and specifi city ( . 95%) Initially, mild elevations are common
in MI and stress cardiomyopathy Quantifi es overall muscle damage
BNP Not reliable in determining cardiac
cause for pulmonary congestion
May predict ICU outcomes Not reliable in obese patients
Normal values excludes cardiac disease
Echocardiogram LVD, MI, valve, and pericardial disease vs
possible stress cardiomyopathy
Noninvasive, bedside Limited quality images in intubated
ICU patients Valuable hemodynamic information
Catheterization CAD diagnosis and revascularization Systolic and diastolic LV function
measurement
Invasive
Higher risk due to ICU comorbidities
Contrast renal injury
Cardiac CT scan Excludes CAD Noninvasive Contrast renal injury. Challenging to
perform in critically ill and unstable
patients.
Cardiac MR scan MI, cardiomyopathies and valve disease
by scarring pattern
Noninvasive Challenging to perform in critically ill,
unstable, and intubated patients Best test for LV regional wall motion
abnormalities and RV function
measurement
BNP 5 brain natriuretic peptide; CAD 5 coronary artery disease; LVD 5 left ventricular dysfunction; MI 5 myocardial infarction; MR 5 magnetic
resonance; RV 5 right ventricular. See Table 1 for expansion of other abbreviation.
© 2010 American College of Chest Physicians
by guest on September 8, 2011chestjournal.chestpubs.orgDownloaded from
CHEST / 138 / 1 / JULY, 2010 201www.chestpubs.org
to be a separate pathogenic entity from LV dysfunction
seen in sepsis. We discuss this with other condi-
tions causing global LV dysfunction in the following
section to emphasize the underlying reversible pre-
cipitating conditions.
Diagnosis of Stress Cardiomyopathy
Angina, heart failure, arrhythmias, ECG changes
(ST elevations or T inversions), and mild troponin
elevations are often triggers for performing cardiac
catheterization. Typically a culprit coronary artery
lesion is not evident. 24 - 26 Left ventriculography dem-
onstrates signifi cant reduction in LV function and
symmetric akinesia involving the mid and apical seg-
ments with relative hypercontractility of the cardiac
base ( Fig 1 ).
Published guidelines require angiographic proof
of absence of coronary artery disease (CAD). 34 In
ICU settings, especially where bleeding risk, severe
comorbidity, or terminal illness precludes catheteri-
zation and revascularization, the following may be
useful indicators that the LV dysfunction is due to
stress cardiomyopathy and not true ACS:
Severe acute LV dysfunction without a signif-1.
icant serum troponin and creatine kinase-MB
elevation.
Symmetrical mid and apical RWMA by echocar-2.
diography—akinesia extending equally in the
inferior and lateral walls as the anteroseptum.
In ACS of the left anterior descending coronary
artery, anteroseptal extent of RWMA from the
apex is usually greater than the inferior and lat-
eral walls. Conversely, ACS of the right coro-
nary or left circumfl ex, if left dominant, usually
spares the anteroseptum.
Repeat echocardiography in a few days to weeks 3.
confi rming complete recovery of LV function
with normalization of typical apical RWMA (in
the absence of lytics and percutaneous coronary
intervention).
Atypical forms of stress cardiomyopathy with RWMA
involving the base, entire LV, or focal LV areas may
be more diffi cult to identify. Table 3 summarizes
salient features of the most common subgroups of LV
dysfunction in critically ill patients. Being noninva-
sive and easy to perform, CT angiography can poten-
tially replace cardiac catheterization in excluding sig-
nifi cant CAD especially when the suspicion for CAD
is low. 14 However, excellent quality coronary imaging
requires patient cooperation (breath holding) and the
ability to tolerate b -adrenergic blockade to slow the
resting heart rate.
Treatment Options for Stress Cardiomyopathy
In ICU settings, identifying and effectively treat-
ing the medical or surgical condition that precipi-
tated stress cardiomyopathy is essential. Supportive
treatment includes addressing heart failure and
arrhythmias as well as optimizing hemodynamics
and metabolic parameters. The prognosis may not
be benign, with one review estimating shock (6.5%),
LV thrombus formation (3.8%), congestive heart
failure (3.8%), and death (3.2%) in patients with stress
cardiomyopathy. 35 There are no randomized data on
stress cardiomyopathy to guide therapy, but prophy-
lactic anticoagulation with warfarin appears reason-
able until LV function recovers. 34 , 35 Because central
sympathetic neurohumoral excess occurs in stress
cardiomyopathy, 30 we believe b -adrenergic block-
ade may be cardioprotective. Because stress car-
diomyopathy is a completely reversible condition,
if the patient can be suffi ciently supported through
the acute phase, long-term cardiac prognosis is gen-
erally good. 34
LV Outfl ow Tract Obstruction in Stress
Cardiomyopathy
About 25% of patients