Increased blood troponin levels in ICU patients

C

Increased blood troponin levels
in ICU patientsNikolaos Markoua, Leonidas Gregorakosb and Pavlos Myrianthefsb
aICU, ‘Thriasio’ Hospital and bAthens University Schoolof Nursing, ICU at «Agioi Anargyroi» General Hospital,Athens, Greece

Correspondence to Nikolaos Markou, 24 Sholiou St,15342 Agia Paraskevi, GreeceTel: +30 6973404627;e-mail: [email protected]

Current Opinion in Critical Care 2011,17:454–463

Purpose of review

Cardiac troponins in serum have become the biomarkers of choice for the diagnosis of

myocardial infarction. Yet, troponin can also be elevated by a multitude of causes in the

absence of overt myocardial ischemia. Such nonspecific elevations are particularly

common in the critically ill. This article aims to provide information on the significance of

troponin elevations in unselected critically ill patients and in patients with sepsis or

septic shock.

Recent findings

Recent studies reconsider the possible association of troponin elevations with

myocardial infarction in unselected critically ill patients. There are also more data on the

prognostic significance of troponin in this population. In patients with sepsis, recent

studies suggest that troponin may be a reliable index of sepsis-induced myocardial

dysfunction, although the implications of this finding in the management of sepsis

remain unclear for the time being. Troponin also appears to be a predictor of short and

possibly also long-term outcome in septic patients. The advent of newer assays with

even higher sensitivity, may lead to further redefinition of the role of troponin in the ICU.

Summary

Troponin is frequently elevated in critically ill patients. More research is needed on the

diagnostic and prognostic significance and possible clinical applications of troponin

measurements in patients with sepsis and critical illness.

Keywords

critical illness, diastolic dysfunction, myocardial infarction, prognosis, sepsis, systolic

dysfunction, troponin

Curr Opin Crit Care 17:454–463� 2011 Wolters Kluwer Health | Lippincott Williams & Wilkins1070-5295

Introduction

Serum levels of cardiac troponins T and I (cTnT, cTnI)

represent an exquisite marker of myocardial injury, with

high myocardial tissue specificity and high sensitivity.

Thus troponins have succeeded creatine phosphokinase-

MB (CPK-MB) as the preferred biomarkers for the

diagnosis of myocardial infarction (MI) and for risk

stratification in acute coronary syndromes (ACS) [1–3].

Currently, for types 1 and 2 of MI, diagnosis requires a

changing pattern of cardiac troponin (a rise or fall within

hours), with at least one value above the upper reference

limit of the assay [1,2,4]. With sensitive assays now in use,

up to 80% of patients presenting with MI can be identified

within 2–3 h, whereas exclusion of MI can take as little as

6 h [5]. The recent development of newer assays with even

lower cut-off points, promises further improvement in the

early diagnosis and risk stratification of acute MI [6,7].

Yet, an elevated troponin does not automatically

equate with diagnosis of MI as it may also be associated

with a multitude of nonischemic causes (given below)

[2,3,8,9].

opyright © Lippincott Williams & Wilkins. Unauth

1070-5295 � 2011 Wolters Kluwer Health | Lippincott Williams & Wilkins

Elevations of troponin in the absence of overt ischemic

heart disease:

(1) C

oriz

ardiac contusion.

(2) A
blation, pacing, cardioversion, endomyocardial
biopsy.

(3) C
ongestive heart failure (acute, chronic).
(4) A
ortic dissection, aortic valve disease or hyper-
trophic cardiomyopathy.

(5) T
achyarrhythmias or bradyarrhythmias.
(6) A
pical ballooning syndrome.
(7) R
habdomyolysis.
(8) P
ulmonary embolism.
(9) S
evere pulmonary hypertension.
(10) R
enal failure.
(11) A
cute neurological disease (stroke, subarachnoid
hemorrhage).

(12) I
nfiltrative myocardial diseases: amyloidosis, hemo-
chromatosis, sarcoidosis, scleroderma.

(13) M
yocarditis or myocardial extension of pericarditis
or endocarditis.

(14) D
rug toxicity: adriamycin, 5-fluorouracil, herceptin,
snake venoms.

ed reproduction of this article is prohibited.

DOI:10.1097/MCC.0b013e3283491f0d

mailto:[email protected]

http://dx.doi.org/10.1097/MCC.0b013e3283491f0d

Increased blood troponin levels in ICU patients Markou et al. 455

(15) C

Cop

Key points
ritically ill patients especially with respiratory fail-
ure or sepsis.
� Troponin elevation is common in critically ill (16) B
patients and is not specific for myocardial ischemia.
urns, especially more than 30% of body surface
area (BSA).
� In these patients, regardless of the presence of (17) E xtreme exertion.
myocardial ischemia, elevation of troponin seems

to be associated with both short- and long-term

prognosis.

� In patients with sepsis, elevations of troponin seem

to be associated with reversible sepsis-induced

myocardial infarction.

It should be noted that in most of these cases troponin

elevation probably still signifies myocardial injury (but

not necessarily necrosis) [4]. Thus diagnosis of myo-

cardial necrosis of ischemic cause in a patient with

increased troponin depends on the clinical context

(symptoms of ischemia, ECG findings, imaging) [1,2].

Increased troponin levels in the absence of ACS should

prompt an evaluation for an alternative, nonthrombotic

mechanism of troponin elevation and direct management

at the underlying cause [6].

A recent study highlights that troponin elevations may

often be unrelated to MI even in acute patients present-

ing with increased clinical likelihood for MI: among 991

consecutive troponin-positive patients presenting with a

clinical suspicion of ACS to the emergency department,

83% of the troponin elevations were retrospectively

attributed to MI, 7.9% were related to other cardiac

causes, and 9.1% to noncardiac diseases. The leading

noncardiac causes were pulmonary embolism, renal fail-

ure, pneumonia, and sepsis [10�]. The specificity for MI

of low-level troponin elevations (one to five times the

upper limit of normal) is probably much lower (�49%)

[11�]. It seems that with sensitive troponin assays cur-

rently in use such low-level elevations represent a gray

zone, when evaluation of the clinical context as well as of

changes of troponin over time becomes even more

imperative for correct diagnosis.

It is well recognized that the levels of troponin are of

prognostic significance in ACS [4]. This prognostic sig-

nificance seems also to extend to patients with acute

noncoronary events (e.g. pulmonary embolism) and to

patients with chronic heart failure and chronic renal

failure [4,9]. Interestingly, in the study of Ilva et al.[10�] noncardiac patients with elevated troponin I at

admission showed significantly higher in-hospital

mortality (26.7 vs. 13.4%) compared to cardiac patients.

The use of recent high-sensitivity assays with very low

cut-offs may aggravate the problem of specificity for MI,

but on the contrary it may offer valuable prognostic

information. It has been recently shown that most

patients with stable coronary artery disease and preserved

left-ventricular function have detectable levels of tropo-

nin with a high-sensitivity troponin T assay, whereas

elevated values are found in 11.1% of patients. In these

patients, troponin concentrations were independently

associated with long-term incidence of cardiovascular

death and heart failure. This association persisted even

yright © Lippincott Williams & Wilkins. Unaut

for values of troponin within normal range of the assay

and below the limit of detection of conventional cardiac

troponin T assays [12].

Cardiac troponin elevation is common in the ICU and can

be observed in up to 40–50% of critically ill patients.

The interpretation, clinical significance and appropriate

management of an elevated troponin measurement in

critically ill patients are uncertain. In these patients,

the pretest probability for ACS is much lower than in

coronary care, whereas nonischemic causes of troponin

elevation (shock, sepsis, pulmonary embolism, renal

failure) are very often encountered. Regardless of cause,

elevated levels of troponin are probably of prognostic

value in critically ill patients, although such associations

are less unequivocal than in ACS [13,14].

We will review the significance of troponin elevations in

the general critically ill patient, with particular emphasis

in more recent findings. We will not discuss the role of

troponin in trauma, cardiac and noncardiac surgery, the

postsurgical setting, revascularization processes or in the

coronary care unit, neither its significance in entities

such as stroke, pulmonary embolism, chronic obstructive

pulmonary disease (COPD), decompensated heart failure

or renal failure.

Troponin in mixed critical care patientsTroponin is commonly increased in critically ill patients,

although the exact frequency varies considerably

between studies, because of differences in patient selec-

tion, case-mix, heterogeneity of assays and different

cut-off values [15–31,32��,33��]. A few years ago, a

meta-analysis of 20 studies with a total of 3278 critically

ill patients, reported that elevated troponin was present

with a median frequency of 43% [interquartile range

(IQR) 21–59%]. Elevated cardiac troponin (cTn) levels

were more frequently found among patients admitted

with sepsis or septic shock. The median frequency of

elevated troponin measurement was higher among

medical ICU patients, somewhat lower in mixed

medical-surgical ICU patients and even lower in surgi-

cal-trauma ICU patients [15]. It should be noted that

horized reproduction of this article is prohibited.

C

456 Cardiovascular system

many of the studies in this meta-analysis concerned not

only unselected critically ill patients but cohorts with

more specific problems (trauma, sepsis, COPD). Table 1

presents data from studies of troponin in unselected

critically ill patients [16–31,32��,33��]. Troponin levels

seem to be higher in more severe disease [22,30,34].

Associations have also been reported with age, cardio-

vascular risk factors and previous cardiovascular history

[30].

Troponin elevation in these patients is not necessarily

indicative of ACS, as many of the entities listed above are

also commonly observed in the ICU. On the other hand,

aggravation of pre-existing coronary artery disease in the

context of extreme stress associated with critical illness

can also lead to increases in troponin because of MI [35].

It has been observed that among patients admitted

to the ICU for gastrointestinal bleeding, those with risk

factors for ischemic heart disease developed ACS more

frequently than those without [36]. It can be easily

postulated that an imbalance in pro and anticoagulant

mechanisms can predispose to thrombotic (type 1) MI

in these patients, whereas endothelial dysfunction might

also play a role in limiting coronary blood flow in patients

with critical coronary lesions. In addition, tachycardia,

hypoxemia, diminished oxygen delivery, may tip the

balance of regional myocardial oxygen consumption over

the critical edge. Thus, in some cases, troponin elevation

might represent not only plaque rupture (type 1 MI) but

also a type 2 MI due to underlying coronary artery disease

and increased myocardial oxygen demand [27,28,35].

Nevertheless, the exact contribution of MI or ACS to

troponin increase in unselected critically ill patients

remains debatable.

In an older prospective study by Ammann et al. [22] in

58 consecutive patients admitted to the ICU for reasons

other than ACS, flow-limiting coronary artery disease

could be excluded with stress echocardiography or

autopsy in 72% of troponin-positive patients. On the

contrary, studies with prospective screening of critically

ill patients with 12-lead electrocardiography (ECG)

and troponin measurements conclude that troponin

elevations are associated with events of myocardial

ischemia in approximately 50% of cases [20,31,32��].

Booker et al. [20] prospectively screened 76 consecutive

patients admitted to a general ICU with 12-lead ECGs for

the first 24–48 h of admission and troponin assays 8–12 h

after the ECG monitoring period. ECG-defined ischemic

events together with troponin I elevation were observed

in six patients, who accounted for 50% of all troponin-

positive results [20].

Lim et al. [31,32��] applied on all patients admitted to a

general ICU over a 2-month period a systematic screen-


ing protocol (with troponin T and 12-lead ECG) over the

whole duration of hospitalization. Patients were screened

at ICU admission and then once daily for the first week;

on alternate days for up to 1 month; then weekly until

ICU death or discharge or for a maximum of 2 months.

Two adjudicators retrospectively reviewed patient charts

to determine the likely cause of troponin elevation. It

should be noted that in their institution patients with

primary cardiac diagnoses but requiring mechanical

ventilation or inotropes/vasopressors were admitted to

the ICU and not to the coronary care unit and thus were

not excluded from the study. Cardiovascular diagnosis

was present in 13.8% of patients but only one of nine

patients with cardiovascular diagnosis had an admission

diagnosis of acute MI, whereas the other eight were

surgical patients.

Of 103 patient admissions, 52 (50.5%) had one or more

elevated troponin measurements. Troponin elevation

was attributed to MI in 53.1% of patients, sepsis in

18.4%, renal failure in 12.2%, and to other causes in

16.3% (cardiac contusion/cardiopulmonary resuscitation

in 6.1%, COPD in 6.1%, congestive heart failure in 4.1%).

Of patients with MI, 61.5% were adjudicated as having

type 1 MI (plaque rupture) on the basis of a history of

coronary artery disease with typical ECG evidence of

progression of infarction (Q waves). A 38.5% of patients

were adjudicated as having type 2 MI (no risk factors or

history of coronary artery disease and presence of a

condition possibly associated with supply–demand

imbalance in the coronary circulation – for example

hypotension, hypovolemia, supraventricular tachycardia).

Interestingly, no significant difference in hospital

mortality was observed between the two types of MI

(25 vs. 40%), although this lack of significance may be

associated with limited number of events [32��].

Diagnosis of MI in these prospective studies was based

on current guidelines, recommending a rise and/or fall in

serum levels of troponin, together with evidence of

myocardial ischemia (clinical, electrocardiographical or

by imaging of new regional wall motion abnormalities)

[2,3]. Yet, diagnosing MI in unselected critically ill

patients remains a challenge. In these patients, not only

is chest pain a rare finding and the significance of troponin

elevation unclear, but interpretation of 12-lead ECG can

be problematic, as even ST-segment elevation has

been reported to be a nonspecific finding that is fre-

quently the result of a variety of nonischemic processes

[20,37,38,39�].

Rennyson et al. [39�] retrospectively studied all cases with

electrocardiographic ST-segment elevation in an ICU: in

23 patients with ST elevation and borderline elevation of

troponin, who should have been classified as MI according

to current guidelines, echocardiographic evaluation could

orized reproduction of this article is prohibited.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.


Ta

ble

1T

rop

on

inin

un

se

lecte

dcri

tica

lly

ill

pa

tie

nts

Ref

eren

ceS

tud

yd

esig

nN

umb

eran

dty

pe

of

pat

ient

scT

nty

pe

Sam

plin

gp

roto

col

Fre

que

ncy

of

elev

ated

cTn

(%)

Ass

oci

atio

no

fcT

nw

ithm

ort

ality

LOS

(una

dju

sted

anal

ysis

)

Gue

stet

al.

[16

]P

rosp

ectiv

e,co

nsec

utiv

e2

09

Med

ical

ICU

cTnI

Dai

ly1

5IC

Um

ort

ality

incr

ease

d(u

niva

riate

anal

ysis

)IC

ULO

Sin

crea

sed

Ko

llef

etal

.[1

7]

Pro

spec

tive,

cons

ecut

ive

26

0M

edic

alIC

UcT

nID

aily

16

Ho

spita

lm

ort

ality

–no

effe

ctin

mul

tivar

iate

anal

ysis

No

ble

etal

.[1

8]

Pro

spec

tive,

case

cont

rol,

cons

ecut

ive

10

9C

ases

and

58

cont

rols

(hea

lthy

volu

ntee

rs)

med

ical

-sur

gic

alIC

U

cTnI

Dai

ly7

1(0

inco

ntro

ls)

Ho

spita

lm

ort

ality

–no

effe

ct(u

niva

riate

anal

ysis

)

Kle

inG

unne

wie

kan

dva

nd

eLe

ur[1

9]

Pro

spec

tive,

cons

ecut

ive

34

Sur

gic

alcT

nTA

dm

issi

on,

next

mo

rnin

gan

d2

4h

late

r3

2N

oas

soci

atio

nw

ithm

ort

ality

No

asso

ciat

ion

with

LOS

Bo

oke

ret

al.

[20

]P

rosp

ectiv

e,co

nsec

utiv

e7

6M

edic

al-s

urg

ical

cTnI

8–

12

haf

ter

24

–4

8h

of

EK

Gm

oni

torin

g1

6A

sso

ciat

ion

with

hosp

ital

mo

rtal

ity(u

nad

just

edan

alys

is)

Rel

os

etal

.[2

1]

Ret

rosp

ectiv

e,no

nco

nsec

utiv

e8

69

Sur

gic

alIC

UcT

nIN

ot

spec

ified

No

tsp

ecifi

edIn

crea

sed

ICU

mo

rtal

ityin

unad

just

edan

alys

isA

mm

ann

etal

.[2

2]

Pro

spec

tive,

cons

ecut

ive.

CA

Dex

clud

edw

ithst

ress

EC

HO

or

auto

psy

58

Med

ical

ICU

cTnI

,cT

nTE

nro

llmen

t,3

,1

2,

24

,4

6,

96

,1

92

h5

5A

sso

ciat

ion

with

ICU

,ho

spita

lm

ort

ality

(uni

varia

tean

alys

is)

Wu

etal

.[2

3]

Pro

spec

tive,

cons

ecut

ive

10

8M

edic

alIC

UcT

nIA

dm

issi

on

45

Incr

ease

dIC

Um

ort

ality

,In

crea

sed

hosp

ital

mo

rtal

ityIn

crea

sed

ICU

LOS

Que

not

etal

.[2

4]

Pro

spec

tive,

cons

ecut

ive

21

7M

edic

alIC

UcT

nIA

dm

issi

on

32

Incr

ease

dho

spita

lm

ort

ality

Incr

ease

dIC

Uan

dho

spita

lLO

SLi

met

al.

[25

]P

rosp

ectiv

e,no

nco

nsec

utiv

e(8

1%

of

all

ICU

adm

issi

ons

)9

3M

edic

al-s

urg

ical

ICU

cTnT

Day

of

adm

issi

on

47

No

tin

dep

end

ent

pre

dic

tor

of

ICU

or

hosp

ital

mo

rtal

ityM

inki

net

al.

[26

]R

etro

spec

tive,

nonc

ons

ecut

ive

13

2/2

21

13

2M

edic

alIC

UcT

nIA

dm

issi

on

31

Una

dju

sted

asso

ciat

ion

with

ICU

but

not

hosp

ital

mo

rtal

ityK

ing

etal

.[2

7]

Pro

spec

tive,

cons

ecut

ive

21

5M

edic

alIC

UcT

nIA

dm

issi

on

(with

in6

h)2

7,

3A

sso

ciat

ion

inun

ivar

iate

anal

ysis

but

not

ind

epen

den

tp

red

icto

ro

f2

8-d

aym

ort

ality

No

asso

ciat

ion

with

ICU

LOS

Land

esb

erg

etal

.[2

8]

Pro

spec

tive.

Pat

ient

sw

ithkn

ow

nst

able

CA

Do

rhi

gh

risk

for

CA

D

10

1G

ener

alIC

UcT

nTD

aily

38

Ass

oci

ated

with

1-m

ont

h,6

-mo

nth

and

2-y

ear

mo

rtal

ityin

univ

aria

teb

utno

tm

ultiv

aria

tean

alys

isLi

met

al.

[29

]R

etro

spec

tive.

Pat

ient

sw

ithex

pec

ted

stay

inIC

U>

72

h;no

nco

nsec

utiv

e

Ear

lysa

mp

ling

(<4

8h)

:1

71

;la

tesa

mp

ling

(>4

8h)

:1

36

;m

edic

al-

surg

ical

ICU

cTnI

As

need

edE

arly

sam

plin

g:

42

;la

tesa

mp

ling

:1

2In

univ

aria

tean

alys

is,

asso

ciat

ion

with

ICU

but

not

hosp

ital

mo

rtal

ity.

No

tin

dep

end

ent

pre

dic

tor

of

surv

ival

Bab

uin

etal

.[3

0]

Ret

rosp

ectiv

e,no

nco

nsec

utiv

e9

29

Med

ical

ICU

cTnT

Ad

mis

sio

n6

1,

4In

dep

end

ent

asso

ciat

ion

with

:3

0-d

aym

ort

ality

,in

-ho

spita

lm

ort

ality

,lo

ng-t

erm

mo

rtal

ityLi

met

al.

[31

,32��

]P

rosp

ectiv

e,co

nsec

utiv

e1

03

Med

ical

surg

ical

ICU

14

%ha

dca

rdio

vasc

ular

dia

gno

sis

on

adm

issi

on,

1ha

dM

I

cTnT

Scr

eene

dat

ICU

adm

issi

on

and

then

onc

ed

aily

for

the

first

wee

k;o

nal

tern

ate

day

sfo

rup

to1

mo

nth;

then

wee

kly

until

ICU

dea

tho

rd

isch

arg

efo

ra

max

imum

of

2m

ont

hs

50

.5In

dep

end

ent

pre

dic

tor

of

hosp

ital

mo

rtal

ityb

utno

to

fIC

Um

ort

ality

Vas

ileet

al.

[33��

]R

etro

spec

tive,

nonc

ons

ecut

ive

20

78

Med

ical

ICU

cTnT

Ad

mis

sio

n4

2,

3In

dep

end

ent

asso

ciat

ion

with

:3

0-d

aym

ort

ality

,in

-ho

spita

lm

ort

ality

,lo

ng-t

erm

mo

rtal

ity

LOS

,le

ngth

of

stay

.D

ata

fro

m[1

6–

31

,32��

,33��

].

C


detect regional or global wall motion abnormalities in only

four cases. Given the fact that the prevalence of both

troponin elevation and of ECG abnormalities in the

ICU is high and that both findings are relatively non-

specific for MI, there is high probability that at least some

patients in the cohort of Lim et al. [32��] with both elevated

troponin and ECG abnormalities did not actually have MI.

Unluckily, neither Booker et al. nor Lim et al. routinely

performed echocardiograms as part of their screening

protocol, whereas angiography was also not part of the

protocol. Thus an unequivocal determination of the exact

frequency of MI in the ICU setting is not possible

at present.

Lim et al. [31] report that in less than half of their MI

patients was a clinical diagnosis possible by ICU staff and

only screening provided 62.2% of MIs ultimately diag-

nosed. They conclude that MI in the ICU can be an

elusive diagnosis in the absence of systematic screening

with troponin and ECGs. An alternative explanation

might be that their screening protocol resulted in over-

diagnosis of MI. Yet, if the findings of Lim et al. are

confirmed by further studies, and the prevalence of MI in

unselected critically ill patients is indeed as high as that

reported by these investigators, then one can ask whether

all critically ill patients should undergo screening (with

troponin and ECG or imaging) for MI. For the time

being, the answer is probably negative: it is currently

unknown whether higher detection rates of MI will

translate into better clinical outcomes through targeted

therapy. The results of the study by Lim et al. also do not

support a strategy of systematic screening: there was

no difference in outcome between patients with MI

detected by staff and MI detected solely by screening.

With clinicians blinded as to the screening diagnosis

which did not affect patient management in their cohort,

this casts doubt to the hypothesis that a diagnosis of MI

can improve outcome in the critically ill [31]. On the

contrary, further prospective studies with screening

(including troponin, ECG and probably echocardiogra-

phy as well) to further clarify the cause of troponin

elevations in the ICU would be highly welcome.

The prognostic significance of troponin inunselected critically ill patientsThere is no unanimity on the prognostic significance

of troponin in unselected critically ill patients [16–31,

32��,33��] (Table 1). Some investigators report that

troponin is an independent predictor of ICU or hospital

mortality [24,30,31,33��] and others find no association

whatever [18,19]. Other studies report associations with

mortality, but they either do not control for other vari-

ables [16,20–23,26] or these associations do not persist in

multivariate analysis [17,25,27].


Associations have also been reported with length of ICU

[16,21,24] and hospital stay [24,40], although in this case

there is no adjustment for other variables.

A meta-analysis concluded that troponin was a univariate

predictor of death in the ICU (eight studies with no

adjustment for confounding variables, with a total of

1019 patients), whereas cumulative data from six more

studies with adjustment for confounding variables (a total

of 1706 patients) demonstrated that troponin was an

independent predictor of death [odds ratio (OR) 2.5;

95% confidence interval (CI) 1.9–3.4]. Moreover, this

meta-analysis concluded that troponin was associated

with an increased length of ICU stay of 3.0 days and

an increased length of hospital stay of 2.2 days (unad-

justed data) [15].

Reaffirmation of a prognostic significance for troponin is

provided by a large retrospective study by Babuin et al.[30]. They reviewed the files of 1657 consecutive unse-

lected patients admitted to medical ICU, 929 of whom

had a cTnT measurement within 6 h of admission.

Cardiovascular causes of admission were not excluded.

Troponin was elevated in 61.3% of the patients, with

mostly small to modest increases. Troponin was an inde-

pendent predictor of 30-day and in-hospital mortality

with no difference observed between cardiovascular

and noncardiovascular admissions.

A recent retrospective study of 240 critically ill patients

with normal (<0.1 ng/ml) or intermediate (0.1–1.49 ng/ml)

cTnI levels, concludes that even borderline elevations

of cTnI are independently associated with in-hospital

mortality and length of ICU stay (but not length of

hospital stay, readmission rate, or postdischarge mortality

at 6 months) [40].

Surprisingly, in the study of Babuin et al. [30] elevated

cTnT on admission was independently associated with

long-term mortality even at 3 years of follow-up, with no

difference observed between cardiovascular and noncar-

diovascular admissions. Two more retrospective studies

from the same group [33��,41] support an association of

troponin at admission to the ICU with long-term out-

come. Elevated troponin upon admission was indepen-

dently associated with in-hospital, short-term and long-

term mortality (up to 3 years follow-up) in a cohort of

2078 patients who had troponin measurement out of a

total of 4433 consecutive admissions to the ICU for a

respiratory condition [33��]. In another retrospective

cohort of 754 patients who had troponin measurements

out of a total of 1076 consecutive ICU admissions for

acute gastrointestinal bleeding, troponin levels on admis-

sion were independently associated with long-term

mortality (though not with short-term mortality) [41].

These findings suggest that even after hospital discharge



there might be an opportunity to intervene in patients

who were troponin-positive on admission to the ICU, in

an effort to improve long-term outcome.

An older study by Lim et al. suggests that the prognostic

significance of troponin elevations in the ICU may vary

depending on the cause of troponin elevation, for example

myocardial ischemia vs. nonischemic origin: they con-

cluded that MI during ICU stay is an independent pre-

dictor of hospital mortality [odds ratio (OR) 3.2], whereas

an isolated increase in troponin was not independently

predictive of ICU or hospital mortality [25]. Later high-

quality prospective data from the same group [31,32��]

indeed confirmed that prognostic significance depends on

the cause of troponin elevation, but the implications

on prognosis were different, with troponin elevation in

the absence of MI appearing to favor a worse outcome. In

this cohort, ICU mortality was 2.0% for patients with no

troponin elevation, 23.1% in patients with MI, and 39.1%

in patients with troponin elevation not due to MI. Elevated

troponin in the absence of MI was independently associ-

ated with hospital (but not ICU) mortality. Associations

of elevated troponin in the presence of MI with mortality

did not persist in multivariate analysis [31].

Troponin in sepsisTroponin elevation is a common finding in patients with

sepsis [42–57,58��,59��] (Table 2). Troponin levels in

these patients seem to be associated with disease severity

[42–47,50,58��,59��] and presence of shock [47,55,59��],

whereas a study of hospitalized bacteremic patients

also reports associations with kidney function, severity

of the underlying infection, and underlying cardiac

disease [60].

The close association of high-sensitivity troponin T

(hs-TnT) with N-terminal pro b-type natriuretic peptide

(NT-proBNP) suggests myocardial origin of troponin

elevation in sepsis [59��]. Although some septic patients

with elevated troponin may have either nonspecific ECG

changes [46,50] or regional wall motion abnormalities

[47], on the whole, neither objective testing [22,45,46]

nor the high occurrence of elevated troponin in a popu-

lation of pediatric sepsis [48] supports the concept that

flow-limiting coronary artery disease is the main cause of

troponin release in sepsis. Troponin release in this

population is most probably the result of low-grade

cytokine-mediated cardiomyocyte injury with transient

loss in membrane integrity and troponin leakage. Sepsis-

related factors like apoptosis, increase in intracellular

calcium, oxidative stress or uncoupling of oxidative phos-

phorylation may also contribute, whereas microvascular

thrombotic injury or systemic hypotension resulting in

suboptimal coronary artery blood flow cannot be excluded

as additive factors in some cases [61–63,64�].

Copyright © Lippincott Williams & Wilkins. Unaut

A recent study concludes that thrombus-associated myo-

cardial damage is unlikely to have a major role in troponin

release in sepsis. The authors studied 38 consecutive

patients with sepsis without evidence of ACS, 58% of

whom were cTnI-positive. Extensive investigation of

coagulation parameters could not demonstrate differences

between cTnI-positive and negative patients in the extrin-

sic, intrinsic or common pathway of coagulation [64�].

It has long been observed that troponin may be associated

with nonischemic systolic dysfunction, mainly in patients

with sepsis [43,45,47,49] but also in unselected critically ill

patients [22]. The more recent findings of Bouhemad et al.[54,65] expand this association to nonischemic diastolic

dysfunction as well, with troponin featuring as a reliable

index of sepsis-induced myocardial dysfunction.

Bouhemad et al. studied prospectively 54 patients with

septic shock: at days 1, 2, 3, 4, 7, and 10 after onset of

septic shock they performed echocardiography with

tissue Doppler and measurement of cTnI and cytokines.

An increase in cTnI was observed in 22 patients, half of

whom had both systolic and diastolic dysfunction

together with acute and left-ventricular dilation, whereas

the rest had isolated impairment of left-ventricular re-

laxation. Echocardiographic findings in patients with

increased cTnI were reversible with reversal of shock.

Improvement of diastolic dysfunction went in parallel with

a progressive decrease in cTnI values and with decreases in

tissue necrosis factor-alpha (TNF-a), IL-8, and IL-10.

The observed improvement in impaired myocardial per-

formance after disease recovery, speaks against major

myocardial cell death in sepsis. On the contrary, patients

with normal cTnI values had either normal systolic and

diastolic function, or in some cases, diastolic dysfunction

that persisted after reversal of shock (and most probably

pre-existed) [54,65]. Associations with mortality were not

assessed in this cohort [54,65], but according to other

investigators diastolic dysfunction is an independent

predictor of hospital survival in septic shock [57].

Although measurements of troponin may facilitate recog-

nition of sepsis-induced myocardial dysfunction, the lack

of established treatment for this entity limits the role of

troponin in the management of sepsis at this time. On the

contrary, troponin can be utilized in the context of

future therapeutic trials in order to target patients with

possible sepsis-induced myocardial dysfunction.

A single-center study by John et al. [50] suggests that

troponin can be used to guide treatment with activated

protein C (APC). In this study patients were prospec-

tively defined but cTnI was measured at the discretion of

the primary physician. The same study group in a retro-

spective analysis of data from the PROWESS study could

not replicate this finding [58��]. Further study is needed


Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.


Ta

ble

2T

rop

on

inin

se

psis

Ref

eren

ceS

tud

yd

esig

nN

umb

eran

dty

pe

of

pat

ient

scT

nty

pe

Sam

plin

gp

roto

col

Fre

que

ncy

of

elev

ated

cTn

(%)

Ass

oci

atio

nw

ithm

ort

ality

Sp

ies

etal

.[4

2]

Pro

spec

tive.

Enr

ollm

ent

not

spec

ified

26

-sur

gic

alIC

UcT

nTE

very

4h

on

day

1.

The

nd

aily

for

7d

ays

69

No

tad

dre

ssed

Fer

nand

eset

al.

[43

]P

rosp

ectiv

e.E

nro

llmen

tno

tsp

ecifi

ed1

0-m

edic

alsu

rgic

alcT

nIW

ithin

12

hfr

om

adm

issi

on

60

Incr

ease

dse

psi

sm

ort

ality

(uni

varia

te)

Tur

ner

etal

.[4

4]

Pro

spec

tive,

cons

ecut

ive.

Cas

e-co

ntro

lM

edic

al-s

urg

ical

;1

5p

atie

nts

with

sep

tican

d6

nons

eptic

sho

ckco

ntro

ls

cTnI

Dai

lyun

tilen

do

fse

psi

s8

0vs

.1

7in

cont

rols

No

tad

dre

ssed

ver

Els

tet

al.

[45

]P

rosp

ectiv

e,co

nsec

utiv

eM

edic

al-s

urg

ical

.4

6p

atie

nts

with

sep

ticsh

ock

cTnI

,cT

nTE

nro

llmen

t,2

4an

d4

8h

50

cTnI

,3

6cT

nTA

sso

ciat

ion

with

hosp

ital

mo

rtal

ity(u

niva

riate

anal

ysis

)A

mm

ann

etal

.[4

6]

Pro

spec

tive,

cons

ecut

ive.

Cas

e-co

ntro

l2

0p

atie

nts

with

sep

sis,

20

nons

eptic

criti

cally

illco

ntro

lscT

nIW

ithin

24

fro

mo

nset

85

vs.

0in

cont

rols

No

tad

dre

ssed

Meh

taet

al.

[47

]P

rosp

ectiv

e,co

nsec

utiv

eM

edic

alsu

rgic

al.

37

sep

ticsh

ock

cTnI

Enr

ollm

ent,

24

,4

8h

43

Incr

ease

dIC

Um

ort

ality

(uni

varia

te)

Fen

ton

etal

.[4

8]

Pro

spec

tive.

Enr

ollm

ent

not

spec

ified

23

child

ren

with

sep

ticsh

ock

cTnI

Initi

al7

2h

57

Briv

etet

al.

[49

]P

rosp

ectiv

e;en

rollm

ent

not

spec

ified

No

tsp

ecifi

ed.

11

8cT

nIN

ot

spec

ified

No

tsp

ecifi

edN

oas

soci

atio

nw

ithho

spita

lm

ort

ality

(mul

tivar

iate

anal

ysis

)Jo

hnet

al.

[50

]P

rosp

ectiv

e;no

nco

nsec

utiv

e1

05

pat

ient

sw

ithse

vere

sep

sis

cTnI

Enr

ollm

ent

46

Incr

ease

dse

psi

sm

ort

ality

(uni

varia

te)

Sco

ttet

al.

[51

]P

rosp

ectiv

e.C

ons

ecut

ive

pat

ient

sre

qui

ring

pul

mo

nary

arte

ryca

thet

er6

6;

surg

ical

ICU

.S

ever

ese

psi

s,se

ptic

sho

ckcT

nIA

tca

thet

eriz

atio

nan

dev

ery

6–

8h

ther

eaft

er6

4N

oas

soci

atio

nw

ithm

ort

ality

Yuc

elet

al.

[52

]P

rosp

ectiv

e,co

nsec

utiv

e4

0p

atie

nts

adm

itted

with

sep

sis

tom

edic

al-s

urg

ical

ICU

cTnI

At

adm

issi

on,

day

2,

day

of

dis

char

ge

No

tsp

ecifi

edA

sso

ciat

ion

(uni

varia

tean

alys

is)

afte

rd

ay1

with

ICU

mo

rtal

ityC

hoo

n-ng

arm

and

Par

tpis

anu

[53

]4

0w

ithse

ptic

sho

ckcT

nT4

2,

5A

sso

ciat

ion

with

mo

rtal

ity(u

niva

riate

anal

ysis

)B

ouh

emad

etal

.[5

4]

Pro

spec

tive.

Enr

ollm

ent

not

spec

ified

Sur

gic

alIC

U.

Sep

ticsh

ock

54

cTnI

.D

ays

1,2

,3

,4

,7,

10

40

Incr

ease

d(u

niva

riate

anal

ysis

)O

livei

raet

al.

[55

]P

rosp

ectiv

e.C

ons

ecut

ive

21

8P

edia

tric

ICU

cTnI

With

in2

4h

fro

men

rollm

ent

4.5

No

asso

ciat

ion

with

hosp

ital

mo

rtal

ity(m

ultiv

aria

tean

alys

is)

Kan

get

al.

[56

]Lo

ng-t

erm

follo

w-u

pfo

rse

psi

so

fa

coho

rto

f3

05

ES

RD

pat

ient

s1

21

ES

RD

pat

ient

sw

ithse

psi

scT

nIO

nset

of

sep

sis

41

.3In

dep

end

ent

asso

ciat

ion

with

3-m

ont

han

d6

-mo

nth

mo

rtal

ityS

turg

ess

etal

.[5

7]

Pro

spec

tive

cons

ecut

ive

Med

ical

-sur

gic

alIC

U2

1se

ptic

sho

ckcT

nTW

ithin

72

ho

fsh

ock

67

No

asso

ciat

ion

with

hosp

ital

mo

rtal

ity(m

ultiv

aria

tean

alys

is)

John

etal

.[5

8��

]R

etro

spec

tive

dat

afr

om

PR

OW

ES

Sst

udy.

No

nco

nsec

utiv

e5

98

of

16

90

pat

ient

s(t

hose

who

had

avai

lab

lere

sults

)cT

nIE

nro

llmen

t7

5In

dep

end

ent

pre

dic

tor

of

28

-day

mo

rtal

ityR

øsj

øet

al.

[59��

]N

onc

ons

ecut

ive;

pat

ient

sad

mitt

edto

ICU

with

seve

rese

psi

sS

ubg

roup

fro

mF

INN

SE

PS

ISst

udy

25

4C

onv

entio

nal

cTnT

assa

y;hs

-cT

nTas

say

Ad

mis

sio

nto

ICU

,7

2h

Ad

mis

sio

n:4

2w

ithfo

urth

-gen

erat

ion

assa

y,8

0w

ithhs

-cT

nIas

say

No

asso

ciat

ion

with

hosp

ital

mo

rtal

ity(m

ultiv

aria

tean

alys

is)

Dat

afr

om

[42

–5

7,5

8��

,59��

].


to determine whether troponin can serve as a simple,

readily available marker to identify which patients with

severe sepsis will benefit from APC, or possibly from

other innovative treatments of sepsis.

Prognostic significance of troponin in septicpatientsSeveral [43,45,47,52], but not all [49,51,55] studies con-

clude that troponin is an independent predictor of

adverse short-term outcome in sepsis (Table 2), whereas

it has also been associated with length of ICU stay in

sepsis patients [47]. Reasons for differences in the pre-

dictive role of troponins in sepsis may be related to small

sample size in some studies or to differences in the

characteristics of troponin assays utilized.

Recently, three sepsis studies reporting on associations

between elevated troponin and outcome were published

[56,58��,59��].

The study of John et al. [58��] was a retrospective analysis

of material from the PROWESS study: baseline cTnI

data were available in 598 nonconsecutive patients from

this cohort. Troponin-positive patients had a significantly

higher 28-day mortality (32 vs. 14%) and this difference

was observed in both the APC-treated and placebo-treated

groups. Elevated cTnI was also an independent predictor

of mortality (OR 2.020; 95% CI 1.153–3.541) [58��].

In a retrospective cohort of 121 patients with end-stage

renal disease (ESRD) and sepsis, cTnI was an indepen-

dent predictor of mortality at 90 days from onset of sepsis

(OR 5.13) and of mortality at more protracted follow-up

(180 days) (hazard ratio 5.90; 95% CI 2.06–16.9) [56].

It should be noted that the cut-off value for troponin

in this study was defined as the value providing the best

diagnostic accuracy for mortality based on receiver

operating characteristic curve (ROC) analysis and not

the cut-off by the manufacturer.

The study by Røsjø et al. [59��] reported on a subgroup of

patients from the prospective FINNSEPSIS study. From

a total of 470 patients who were enrolled with severe

sepsis or septic shock in this cohort, 254 consented for

blood sampling for troponin: cTnT levels were measured

by a novel hs-cTnT assay at two time points (inclusion

and 72 h thereafter). Results for the hs-cTnT assay were

compared to those of the established fourth-generation

cTnT assay (lower limit of detection 0.003 mg/l vs.

0.01 mg/l for fourth-generation assay). At inclusion in

the study, cTnT was detectable in 60% of patients with

the conventional assay and in 100% with the hs-cTnT

assay, whereas increased levels were found in 42% with

the standard assay and in 80% with hs-cTnT assay. At

72 h, troponin was still detected in 57% of patients with

the conventional assay and in 100% with hs-cTnT assay,


whereas it remained elevated in 7% of patients with the

conventional assay and in 79% with the hs-cTnT assay.

Survivors had higher levels of hs-cTnT on inclusion,

whereas no difference in troponin values was observed

with the conventional assay. Troponin was not indepen-

dently associated with in-hospital mortality, regardless of

assay but, interestingly, hs-cTnT on inclusion was inde-

pendently associated with development of shock during

hospitalization (OR 2.45; 95% CI 1.09–5.53) and could

predict development of septic shock with a sensitivity of

86% and specificity of 33% [59��]. This role of cTnT as an

index of impending shock in early sepsis needs to be

confirmed in further studies.

The study by Røsjø et al. highlights the importance of

the characteristics of troponin assays which may be a

partial explanation for the divergence of results between

existing studies of troponin in critical illness and sepsis.

ConclusionTroponin is frequently elevated in the critically ill, and

recent data suggest that MI is the underlying (and often

unrecognized) cause of such elevations in a large pro-

portion of these patients. Recent studies also reinforce

the impression that troponin may be an independent

predictor of short-term outcome in critical illness and

expand the predictive role of troponin to the long-term

outcome of this patient population.

In sepsis, troponin elevations do not seem to be com-

monly related to ACS, but are probably associated with

reversible sepsis-induced myocardial dysfunction. There

are newer data on the predictive role of troponin in sepsis,

although the verdict on this question is not unanimous.

Recent findings on troponin need to be confirmed in

larger and more systematic prospective trials enrolling

consecutive patients. More research is also needed on the

possible implications of these findings in the clinical

management of mixed critically ill patients and patients

with severe sepsis or septic shock.

AcknowledgementsConflicts of interestThere are no conflicts of interest.

References and recommended readingPapers of particular interest, published within the annual period of review, havebeen highlighted as:� of special interest�� of outstanding interest

Additional references related to this topic can also be found in the CurrentWorld Literature section in this issue (p. 539).

1 Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction rede-fined-a consensus document of The Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardialinfarction. J Am Coll Cardiol 2000; 36:959–969.


C


2 Thygesen K, Alpert JS, White HD, Joint ESC/ACCF/AHA/WHF Task Forcefor the Redefinition of Myocardial Infarction. Universal definition of myocardialinfarction. Circulation 2007; 116:2634–2653.

3 Thygesen K, Alpert JS, Jaffe AS, White HD. Diagnostic application of theuniversal definition of myocardial infarction in the intensive care unit. CurrOpin Crit Care 2008; 14:543–548.

4 Bertsch T, Braun SL, Giannitis E, et al. Conference on clinical use of troponinT high sensitive (TnThs) on September 8, 2009 at the airport conferencecenter, Frankfurt/Main. Clin Lab 2010; 56:355–358.

5 Mac Rae AR, Kavsak PA, Lustig V. Assessing the requirement for the 6-hourinterval between specimens in the American Heart Association Classificationof Myocardial Infarction in epidemiology and clinical research studies. ClinChem 2006; 52:812–818.

6 Keller T, Zeller T, Peetz D, et al. Sensitive troponin I assay in early diagnosis ofacute myocardial infarction. N Engl J Med 2009; 361:868–877.

7 Reichlin T, Hochholzer W, Bassetti S, et al. Early diagnosis of myocardialinfarction with sensitive cardiac troponin assays. N Engl J Med 2009;361:858–867.

8 Jeremias A, Gibson CM. Narrative review: alternative causes for elevatedcardiac troponin levels when acute coronary syndromes are excluded. AnnIntern Med 2005; 142:786–791.

9 Agzew Y. Elevated serum cardiac troponin in nonacute coronary syndrome.Clin Cardiol 2009; 32:15–20.

10

�Ilva TJ, Eskola MJ, Nikus KC, et al. The etiology and prognostic significanceof cardiac troponin I elevation in unselected emergency department patients.J Emerg Med 2010; 38:1–5.

Large prospective study of causes of troponin elevation in the emergency depart-ment.

11

�Lee HM, Kerr D, O’H Ici D, Kelly AM. Clinical significance of initial troponin I inthe grey zone in emergency department chest pain patients: a retrospectivepilot study. Emerg Med J 2010; 27:302–304.

A retrospective study of patients assessed for ACS in emergency department.

12 Omland T, de Lemos JA, Sabatine MS, et al. Prevention of Events withAngiotensin Converting Enzyme Inhibition (PEACE) Trial Investigators. Asensitive cardiac troponin T assay in stable coronary artery disease. N EnglJ Med 2009; 361:2538–2547.

13 Fromm RE Jr. Cardiac troponins in the intensive care unit: common causesof increased levels and interpretation. Crit Care Med 2007; 35:584–588.

14 Lim W. JOE DOUPE LECTURE: cardiac troponin in the intensive care unit.Clin Invest Med 2009; 32:E405–E410.

15 Lim W, Qushmaq I, Devereaux PJ, et al. Elevated cardiac troponin measure-ments in critically ill patients. Arch Intern Med 2006; 166:2446–2454.

16 Guest TM, Ramanathan AV, Tuteur PG, et al. Myocardial injury in critically illpatients. A frequently unrecognized complication. J Am Med Assoc 1995;273:1945–1949.

17 Kollef MH, Ladenson JH, Eisenberg PR. Clinically recognized cardiac dys-function: an independent determinant of mortality among critically ill patients.Is there a role for serial measurement of cardiac troponin I? Chest 1997;111:1340–1347.

18 Noble JS, Reid AM, Jordan LV, et al. Troponin I and myocardial injury in theICU. Br J Anaesth 1999; 82:41–46.

19 Klein Gunnewiek JM, van de Leur JJ. Elevated troponin T concentrations incritically ill patients. Intensive Care Med 2003; 29:2317–2322.

20 Booker KJ, Holm K, Drew BJ, et al. Frequency and outcomes of transientmyocardial ischemia in critically ill adults admitted for noncardiac conditions.Am J Crit Care 2003; 12:508–516.

21 Relos RP, Hasinoff IK, Beilman GJ. Moderately elevated serum troponinconcentrations are associated with increased morbidity and mortalityrates in surgical intensive care unit patients. Crit Care Med 2003; 31:2598–2603.

22 Ammann P, Maggiorini M, Bertel O, et al. Troponin as a risk factor for mortalityin critically ill patients without acute coronary syndromes. J Am Coll Cardiol2003; 41:2004–2009.

23 Wu TT, Yuan A, Chen CY, et al. Cardiac troponin I levels are a risk factor formortality and multiple organ failure in noncardiac critically ill patients and havean additive effect to the APACHE II score in outcome prediction. Shock 2004;22:95–101.

24 Quenot JP, Le Teuff G, Quantin C, et al. Myocardial injury in critically illpatients: relation to increased cardiac troponin I and hospital mortality. Chest2005; 128:2758–2764.

25 Lim W, Qushmaq I, Cook DJ, et al., Troponin T Trials Group. Elevated troponinand myocardial infarction in the intensive care unit: a prospective study. CritCare 2005; 9:R636–R644.


26 Minkin R, Cotiga D, Noack S, et al. Use of admission troponin in critically illmedical patients. J Intensive Care Med 2005; 20:334–338.

27 King DA, Codish S, Novack V, et al. The role of cardiac troponin I as aprognosticator in critically ill medical patients: a prospective observationalcohort study. Crit Care 2005; 9:R390–R395.

28 Landesberg G, Vesselov Y, Einav S, et al. Myocardial ischemia, cardiactroponin, and long-term survival of high-cardiac risk critically ill intensive careunit patients. Crit Care Med 2005; 33:1281–1287.

29 Lim W, Cook DJ, Griffith LE, et al. Elevated cardiac troponin levels in criticallyill patients: prevalence, incidence, and outcomes. Am J Crit Care 2006;15:280–288.

30 Babuin L, Vasile VC, Rio Perez JA, et al. Elevated cardiac troponin is anindependent risk factor for short- and long-term mortality in medical intensivecare unit patients. Crit Care Med 2008; 36:759–765.

31 Lim W, Holinski P, Devereaux PJ, et al. Detecting myocardial infarction incritical illness using screening troponin measurements and ECG recordings.Crit Care 2008; 12:R36.

32

��Lim W, Whitlock R, Khera V, et al. Etiology of troponin elevation in critically illpatients. J Crit Care 2010; 25:322–328.

Prospective study applying systematic screening with troponin and ECG inunselected critically ill patients and searching for causes of elevation of troponin.

33

��Vasile VC, Chai HS, Khambatta S, et al. Significance of elevated cardiactroponin T levels in critically ill patients with acute respiratory disease. Am JMed 2010; 123:1049–1058.

Large retrospective study of troponin in patients admitted to the ICU for respiratoryconditions. Associations with both short-term and long-term mortality were ob-served.

34 Arlati S, Brenna S, Prencipe L, et al. Myocardial necrosis in ICU patients withacute noncardiac disease: a prospective study. Intensive Care Med 2000;26:31–37.

35 King DA, Almog Y. Myocardial infarction complicating critical illness. Crit Care2005; 9:634–635.

36 Bhatti N, Amoateng-Adjepong Y, Qamar A, Manthous CA. Myocardialinfarction in critically ill patients presenting with gastrointestinal hemor-rhage: retrospective analysis of risks and outcomes. Chest 1998; 114:1137–1142.

37 Lim W, Qushmaq I, Cook DJ, et al. Reliability of electrocardiogram interpreta-tion in critically ill patients. Crit Care Med 2006; 34:1338–1343.

38 Lim W, Tkaczyk A, Holinski P, et al. The diagnosis of myocardial infarction incritically ill patients: an agreement study. J Crit Care 2009; 24:447–452.

39

�Rennyson SL, Hunt J, Haley MW, et al. Electrocardiographic ST-segmentelevation myocardial infarction in critically ill patients: an observational cohortanalysis. Crit Care Med 2010; 38:2304–2309.

A retrospective study evaluating the significance of ST-segment elevation in theICU, in association with troponin levels.

40 Stein R, Gupta B, Agarwal S, et al. Prognostic implications of normal(<0.10 ng/ml) and borderline (0.10 to 1.49 ng/ml) troponin elevation levelsin critically ill patients without acute coronary syndrome. Am J Cardiol 2008;102:509–512.

41 Vasile VC, Babuin L, Rio Perez JA, et al. Long-term prognostic significance ofelevated cardiac troponin levels in critically ill patients with acute gastro-intestinal bleeding. Crit Care Med 2009; 37:140–147.

42 Spies C, Haude V, Fitzner R, et al. Serum cardiac troponin T as a prognosticmarker in early sepsis. Chest 1998; 113:1055–1063.

43 Fernandes CJ Jr, Akamine N, Knobel E. Cardiac troponin: a new serummarker of myocardial injury in sepsis. Intensive Care Med 1999; 25:1165–1168.

44 Turner A, Tsamitros M, Bellomo R. Myocardial cell injury in septic shock. CritCare Med 1999; 27:1775–1780.

45 ver Elst KM, Spapen HD, Nguyen DN, et al. Cardiac troponins I and T arebiological markers of left ventricular dysfunction in septic shock. Clin Chem2000; 46:650–657.

46 Ammann P, Fehr T, Minder EI, et al. Elevation of troponin I in sepsis and septicshock. Intensive Care Med 2001; 27:965–969.

47 Mehta NJ, Khan IA, Gupta V, et al. Cardiac troponin I predicts myocardialdysfunction and adverse outcome in septic shock. Int J Cardiol 2004; 95:13–17.

48 Fenton KE, Sable CA, Bell MJ, et al. Increases in serum levels of troponin I areassociated with cardiac dysfunction and disease severity in pediatric patientswith septic shock. Pediatr Crit Care Med 2004; 5:533–538.

49 Brivet FG, Jacobs FM, Colin P, et al. Cardiac troponin level is not anindependent predictor of mortality in septic patients requiring medicalintensive care unit admission. Crit Care 2006; 10:404.



50 John J, Awab A, Norman D, et al. Activated protein C improves survival in severesepsis patients with elevated troponin. Intensive Care Med 2007; 33:2122–2128.

51 Scott EC, Ho HC, Yu M, et al. Preexisting cardiac disease, troponin I elevationand mortality in patients with severe sepsis and septic shock. AnaesthIntensive Care 2008; 36:51–59.

52 Yucel T, Memis D, Karamanlioglu B, et al. The prognostic value of atrial andbrain natriuretic peptides, troponin I and C-reactive protein in patients withsepsis. Exp Clin Cardiol 2008; 13:183–188.

53 Choon-ngarm T, Partpisanu P. Serum cardiac troponin-T as a prognosticmarker in septic shock. J Med Assoc Thai 2008; 91:1818–1821.

54 Bouhemad B, Nicolas-Robin A, Arbelot C, et al. Isolated and reversibleimpairment of ventricular relaxation in patients with septic shock. Crit CareMed 2008; 36:766–774.

55 Oliveira NS, Silva VR, Castelo JS, et al. Serum level of cardiac troponin I inpediatric patients with sepsis or septic shock. Pediatr Crit Care Med 2008;9:414–417.

56 Kang EW, Na HJ, Hong SM, et al. Prognostic value of elevated cardiactroponin I in ESRD patients with sepsis. Nephrol Dial Transplant 2009;24:1568–1573.

57 Sturgess DJ, Marwick TH, Joyce C, et al. Prediction of hospital outcome inseptic shock: a prospective comparison of tissue Doppler and cardiacbiomarkers. Crit Care 2010; 14:R44.

58

��John J, Woodward DB, Wang Y, et al. Troponin-I as a prognosticator ofmortality in severe sepsis patients. J Crit Care 2010; 25:270–275.

Retrospective analysis of material from the PROWESS study. Troponin was anindependent predictor of mortality in patients with sepsis.


59

��Røsjø H, Varpula M, Hagve TA, et al., FINNSEPSIS Study Group. Circulatinghigh sensitivity troponin T in severe sepsis and septic shock: distribution,associated factors, and relation to outcome. Intensive Care Med 2011;37:77–85.

Retrospective analysis of data from the prospective FINNSEPSIS study. Compar-ison of a conventional cTnT assay, with a novel hs-cTnT assay in patients withsepsis.

60 Kalla C, Raveh D, Algur N, et al. Incidence and significance of a positivetroponin test in bacteremic patients without acute coronary syndrome. Am JMed 2008; 121:909–915.

61 Maeder M, Fehr T, Rickli H, Ammann P. Sepsis-associated myocardialdysfunction. Diagnostic and prognostic impact of cardiac troponins andnatriuretic peptides. Chest 2006; 129:1349–1366.

62 Fernandes CJ Jr, Akamine N, Knobel E. Myocardial depression in sepsis.Shock 2008 (Suppl 1):S14–S17.

63 Favory R, Neviere R. Significance and interpretation of elevated troponin inseptic patients. Crit Care 2006; 10:224.

64

�Altmann DR, Korte W, Maeder MT, et al. Elevated cardiac troponin I in sepsisand septic shock: no evidence for thrombus associated myocardial necrosis.PLoS One 2010; 5:e9017.

Prospective study investigating the association of troponin elevation in sepsis withcoagulation abnormalities.

65 Bouhemad B, Nicolas-Robin A, Arbelot C, et al. Acute left ventricular dilatationand shock-induced myocardial dysfunction. Crit Care Med 2009; 37:441–447.


Increased blood troponin levels in ICU patients

Documents