IJC Metabolic & Endocrine 10 (2016) 30–35

Contents lists available at ScienceDirect

IJC Metabolic & Endocrine

journal homepage: http://www.journals.elsevier.com/ijc-metabolic-and-endocrine

N-terminal-pro-brain natriuretic peptide elevations in the course of

septic and non-septic shock reflect systolic left ventricular dysfunction
assessed by transpulmonary thermodilution
A.B. Johan Groeneveld a,⁎, Ronald J. Trof b
a
Department of Intensive Care, Erasmus Medical Center, Rotterdam, The Netherlands
b
Medisch Spectrum Twente, Enschede, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: Background: The cardiac correlates, if any, of N-terminal probrain natriuretic peptide (NT-proBNP) levels in septic
Received 5 November 2015 and non-septic shock patients remain controversial.
Received in revised form 26 November 2015 Methods: In the 38 septic and 22 non-septic shock patients in the transpulmonary thermodilution arm of a
Accepted 10 January 2016 previous 2-center randomized controlled trial comparing pulmonary artery catheterization with transpulmonary
Available online 17 January 2016
thermodilution, serial (daily for 3 days) and paired measurements (n = 145) were obtained of NT-proBNP and
transpulmonary dilution variables as global ejection fraction (GEF), left ventricular preload-recruitable stroke
Keywords:
Natriuretic peptides
work (PRSW) and diastolic compliance.
Shock Results: Elevated NT-proBNP inversely related to low GEF and PRSW in pooled data (r = −0.45, P b 0.001). The
Sepsis 72 h course of NT-proBNP was inversely associated with PRSW, independent of age, gender, creatinine, norepi-
Cardiac function nephrine treatment and diastolic compliance, without differences between septic and non-septic shock. Over
Renal function the 72 h study period, NT-proBNP levels were higher in 28 day non-survivors than survivors, independent of
Transpulmonary thermodilution type of shock and disease severity.
Conclusions: In septic and non-septic shock, NT-proBNP elevations reflect systolic left ventricular dysfunction and
are associated with a poor outcome. They may help recognition of cardiac dysfunction in shock and its manage-
ment when invasive hemodynamic monitoring is not yet instituted.
© 2016 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction proBNP even without hemodynamic changes or cardiac dysfunction

[33,37].
The causes and consequences of plasma elevations of the cardiac N- Methods to evaluate cardiac function in the critically ill to elucidate
terminal prohormone of brain natriuretic peptide (NT-proBNP) in the the mechanisms of (NT-pro)BNP release included echocardiography
critically ill remain highly controversial, partly because many factors [1,4,6,8,10,11,14,15,20,24,32,33,36] and pulmonary artery catheteriza-
may be involved [1–3]. Elevated levels have been used, with varying tion [3–5,7,10,13,16,34,38] but many studies did not look at both
success, to help diagnosing the cause of respiratory insufficiency diastolic and systolic function indices nor compared septic and non-
[3–10] and to prognosticate outcome of critical illness after surgery or septic patients with different etiology and severity of cardiac disease
sepsis [9,11–28], among others. In contrast, shock and hypovolemia [1,4,8,11,13,15,20,21,24,26,32,33,35,38]. In any case, the relation
are expected not to increase the hormone levels, because of decreased between filling pressures of the heart and NT-proBNP levels was poor
cardiac wall stress (increased stiffness may be pivotal in releasing natri- at best [3–5,7,9,10,11,25,33]. Many studies used single admission values
uretic peptides [29]), unless associated with cardiac disease or involve- rather than courses over time [1,4,5,15,19,23,32], while confounding by
ment [6,13,30–32]. In sepsis-like conditions, a key issue is whether BNP age, gender, obesity, renal disease, and treatment [1] has not always
or its prohormone is a marker of the inflammatory response [2,12,18,31, been taken into account [2,3,7,15,20,26,33]. Serial measurements have
33,34], or of diastolic or systolic cardiac dysfunction [11,14,23,24,35] or been obtained in only few studies and some addressed the changes in
both [32,36]. Indeed, endotoxin injection or sepsis may elevate NT- cardiac filling and function in time and their relation to BNP levels,
which was sometimes poor [3,24,33,38].
In contrast, we evaluated cardiac function by transpulmonary
⁎ Corresponding author at: Department of Intensive Care, Erasmus Medical Center, 's
thermodilution (TPTD) and suggested that NT-proBNP elevations
Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands. inversely related to, primarily, systolic rather than diastolic function
E-mail address: [email protected] (A.B.J. Groeneveld). parameters, both in sepsis and non-sepsis [39]. The latter study included

http://dx.doi.org/10.1016/j.ijcme.2016.01.002
2214-7624/© 2016 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 A.B.J. Groeneveld, R.J. Trof / IJC Metabolic & Endocrine 10 (2016) 30–35 31

systolic function indices as the global ejection fraction (GEF), ie enrollment. Fluid challenges were performed by synthetic colloids at a
stroke volume divided by global end-diastolic volume, and preload- dose of 250–500 mL per 30 min when indicated clinically and according
recruitable stroke work index (PRSW), represented by the ratio of left to the TPTD parameters defining upper limits of safe infusion. Fluid chal-
ventricular stroke work to end-diastolic volume, which may be less lenges were withheld when the safety limits by TPTD monitoring had
dependent on (after)loading than the GEF [39–41]. As a diastolic func- been reached and when there was ≥ 10% rise in cardiac output.
tion index the end-diastolic volume to pressure ratio (compliance) Norepinephrine was the vasopressor drug of first choice in our ICU's,
can be used [39]. The study [39], however, was relatively small (18 sep- which was continuously infused and dosed on the basis of hemodynam-
sis patients not in shock) and lacked longitudinal measurements to ic responses. End points of resuscitation, reflecting adequacy of hemo-
answer the question whether NT-pro BP levels in the course of time in- dynamic management, were mean arterial pressure ≥ 65 mm Hg (or
deed reflect changes in left ventricular function. ≥80 mm Hg in case of known hypertension), ScvO2 ≥70%, lactate clear-
Because of the continuing controversies, we set out to serially evalu- ance, diuresis ≥ 0.5 mL/kg/h (unless development of intrinsic renal
ate NT-proBNP levels in TPTD-monitored shock patients of septic and failure), and restoration of peripheral perfusion deficits.
non-septic origin. The hypothesis was that the hormone levels better
reflect systolic than diastolic indices of cardiac dysfunction both in sep- 2.3. Therapeutic protocol
tic and non-septic shock.
All patients were pressure-controlled ventilated (Servo-i, Maquette,
2. Patients and methods Sweden or Evita 4, Dräger, Lübeck, Germany), aiming at tidal volumes
b8 mL/kg predicted body weight and positive end-expiratory pressure
This is a randomized, non-blinded, 2-centre clinical trial conducted was dosed (≤20 cm H2O) to maintain arterial PO2 ≥65 mm Hg at an in-
in intensive care units (ICUs) in The Netherlands from February 2007 spiratory O2 fraction of about 40%. Pressure-controlled ventilation was
to July 2009 (trial registration number: NL14119.029.06, CCMO, The changed into pressure-support ventilation when clinically justified.
Netherlands) reported previously [42]. Patients meeting inclusion Weaning was attempted through clinical protocols. After 72 h, vaso-
criteria were randomly assigned to receive either a TPTD or pulmonary pressors and inotropic drugs were administered and dosed on clinical
artery catheter when inclusion criteria for advanced hemodynamic grounds. Sedatives, analgesics, and antibiotics were also prescribed by
monitoring were met. For the current research we only included the attending physicians according to clinical guidelines. Systemic cortico-
TPTD arm of the study and therefore only describe the protocol followed steroids were initialized in case of persistent vasopressor-dependent
in this arm of the study. The medical ethics committee at each study septic shock, defined as a norepinephrine dose N 1 mg/h.
center had approved of the protocol.
2.4. Data collection, measurements and assays
2.1. In- and exclusion criteria
Hemodynamic measurements were done at the mid chest level with
In brief, patients were eligible when they were on mechanical venti- patients in supine position, after calibration and zeroing to atmospheric
lation with an expected stay in the ICU N 48 h in the presence of shock, pressure, at least every 24 h (0, 24, 48 and 72 h after enrolment). These
indicating a clinical reason for invasive hemodynamic monitoring. We included measurements of mean arterial pressure (MAP, mm Hg) via
consecutively included patients in septic and non-septic shock. Shock routine radial artery catheter, central venous pressure (CVP, mm Hg)
was defined by acute circulatory failure characterized by persistent and TPTD parameters. The latter were performed through an injection
arterial hypotension defined as a mean arterial pressure b 65 mm Hg of a 20-mL ice-cold (4 °C) saline bolus through a central venous cathe-
(or b80 mm Hg with previous hypertension) despite assumingly ter. Measurements were obtained in triplicate and averaged, including
adequate volume resuscitation and/or the need for vasopressors to cardiac output and global end-diastolic volume, indexed to body surface
maintain a mean arterial pressure ≥65 mm Hg (or ≥80 mm Hg in case yielding cardiac index (CI) and global end-diastolic volume index
of known hypertension). Septic shock was defined by shock plus two (GEDVI, n 680–800 mL/m2), respectively. Global ejection fraction
or more of the following for systemic inflammatory response syndrome (GEF, n 25–35%) was calculated from stroke volume (CI/heart rate)/
criteria: abnormal body temperature (N38 °C, b36 °C), tachycardia GEDVI/4, a measure of left ventricular ejection fraction in the absence
(N90 beats/min), mechanical ventilation, and abnormal white blood of severe right heart dysfunction [43,44]. Left ventricular stroke work
cell counts (≤4 or ≥12 × 109 /L or N10% immature bands) plus a clinical- (LVSWI, n 45–60 cJ/m2) was calculated from mean arterial pressure
ly evident and/or microbiologically proven focus on infection. Non- times stroke volume index (× 0.0136). PRSW was derived from
septic shock was shock after 1) (surgery for) major trauma (Injury LVSWI/GEDVI/4 (or MAP × GEF, n 20–30 mm Hg), a relatively load-
Severity Score N25, without documented traumatic brain injury); independent measure of left ventricular function in the absence of
2) elective and emergency major abdominal surgery (including esoph- severe right heart dysfunction [40,41], and diastolic compliance from
ageal or gastric resection, liver surgery, pancreatic surgery, colorectal GEDVI/CVP, mL/mm Hg/m2. Arterial blood samples were taken at least
surgery); 3) cardiac surgery (coronary bypass surgery, aortic root and/ every 24 h, at baseline, and up to 72 h after enrollment. Lactate
or valvular surgery); 4) major vascular surgery (aorta and iliac/mesen- (n b 2 mmol/L) and creatinine levels (n b 130 μmol/L) were measured
teric reconstructions); and 5) cardiogenic shock or terminal congestive (i-STAT 1, Abbott, Abbott Park, IL, USA and Boehringer-Mannheim
heart failure. There were no patients with obstructive shock caused by Hitachi analyzers 911 and 747, Almere, The Netherlands). For measure-
pulmonary embolism. Exclusion criteria were age b18 or N80 years, ment of NT-proBNP plasma levels blood was collected in tubes contain-
pregnancy, preterminal illness with life expectancy b24 h, therapeutic ing EDTA. Within 2 h, blood samples were centrifuged for 10 min at
hypothermia after cardiac arrest, traumatic brain injury, known 3000 rpm at ambient temperature. Plasma was stored at −80 °C until
(unrepaired) cardiac or vascular aneurysms, bifemoral vascular surgery analyzed. An electrochemiluminescence immunoassay for NT-proBNP
or known pulmonary hypertension. was performed with the Modular analytics E170 system (Roche, Mann-
heim, Germany). The upper limits of normal for healthy volunteers
2.2. Study protocol (95th percentile) according to the manufacturer's recommendations
for age were 97.3 pg/mL (18–44 years), 121 pg/mL (45–54 years),
In brief, the TPTD catheter (PiCCO; Pulsion Medical Systems AG, 198 pg/mL (55–64 years), 285 pg/mL (65–74 years), and 526 pg/mL
Munich, Germany) was inserted in the femoral artery. Fluid resuscita- (≥75 years). The lower detection limit was 5 pg/mL. The upper detec-
tion and hemodynamic management was guided by TPTD-derived tion limit of the assay for diluted samples was 70,000 pg/mL. Doses
parameters according to a predefined algorithm [42], up to 72 h after and types of inotropic/vasopressor drugs were registered. Daily
32 A.B.J. Groeneveld, R.J. Trof / IJC Metabolic & Endocrine 10 (2016) 30–35

measured clinical and laboratory variables allowed calculation of Se- Table 2

quential Organ Failure Assessment (SOFA score). Patients were followed Hemodynamic and biochemical values at baseline and 72 h.

until death or hospital discharge. Septic shock Non-septic shock P

N = 38 n = 22

HR, /min t=0 107 ± 22 89 ± 19 0.002

2.5. Statistical analysis t = 72 91 ± 22 90 ± 14 0.75
CVP, mm Hg t=0 12 ± 6 12 ± 5 0.92
t = 72 9±4 11 ± 5 0.11
The main study parameters were non-normally distributed
MAP, mm Hg t=0 78 ± 14 83 ± 18 0.30
(Kolmogorov–Smirnov test P b 0.05) and logarithmically converted t = 72 87 ± 12 90 ± 16 0.42
before analysis. Baseline variables were compared with Student-t, GEDVI, mL/m2 t=0 779 ± 174 831 ± 217 0.31
Mann–Whitney U or Fisher exact test, where appropriate. We used gen- t = 72 759 ± 170 901 ± 232 0.041
eralized estimating equations (GEE) to study relations of NT-proBNP Compliance, mL/mm Hg/m2 t=0 76 ± 40 79 ± 33 0.82
t = 72 150 ± 158 106 ± 87 1.0
plasma levels with variables in time, taking repeated observations in CI, L/min/m2 t=0 3.5 ± 1.3 2.9 ± 0.9 0.054
the same patients into account. Standardized regression coefficients t = 72 3.9 ± 1.1 3.4 ± 0.8 0.11
were calculated. Receiver operating characteristic curves (ROC) were LVSWI, cJ/m2 t=0 34 ± 12 37 ± 12 0.47
constructed to evaluate the predictive values of NT-proBNP levels and t = 72 53 ± 16 48 ± 16 0.36
GEF, % t=0 18 ± 7 17 ± 8 0.86
optimal cutoff values (Youden index) with associated sensitivity, speci-
t = 72 25 ± 8 19 ± 6 0.01
ficity and positive and negative predictive values were calculated. All PRSW, mm Hg t=0 14 ± 5 14 ± 6 0.87
tests were two-sided and a P b 0.05 was considered statistically signifi- t = 72 18 ± 7 15 ± 6 0.054
cant. Exact P values N 0.001 are reported. Data are summarized as Norepinephrine, μg/kg/min t=0 0.53 ± 0.39 0.28 ± 0.28 0.027
mean ± standard deviation (SD). t = 72 .22 ± 0.33 0.11 ± 0.18 0.29
Lactate, mmol/L t=0 4.1 ± 3.0 3.4 ± 2.4 0.3
t = 72 2.0 ± 1.2 1.4 ± 0.8 0.06
Creatinine, μmol/L t=0 174 ± 140 174 ± 99 0.99
3. Results t = 72 138 ± 100 170 ± 101 0.27
Leukocytes, ×109/L t=0 13.1 ± 10.5 11.7 ± 10.7 0.54
t = 72 17.5 ± 8.0 17.5 ± 22.0 0.99
3.1. Patient characteristics
NT-proBNP, pg/mL t=0 9028 ± 16,627 3986 ± 5450 0.48
t = 72 3249 ± 5776 4691 ± 5192 0.89
Table 1 summarizes patient data. Disease severity and in hospital
Mean ± SD. Abbreviations: HR, heart rate; CVP, central venous pressure; MAP, mean arte-
mortality was higher in septic than in non-septic shock. Non-septic rial pressure; GEDVI, global end-diastolic volume index; CI, cardiac index; LVSWI, left
shock patients had more cardiac premorbidity. Within the 72 h observa- ventricular stroke work index; GEF, global ejection fraction; PRSW, preload-recruitable
tion period 9 septic patients had died and 2 had been discharged, so stroke work.
that, together with technical failures, 65 TPTD measurements and 68
NT-proBNP measurements at the 240 observation points were missing
and 145 paired observations over time were available for analysis. 3.2. NT-proBNP and cardiac function

In Table 2 the courses of hemodynamic and biochemical variables in

the shock groups are summarized and compared. Septic shock patients
Table 1
Patient characteristics.
had a more increased heart rate, higher CI, lower GEDVI and reversible
depression of GEF and PRSW as compared to non-septic shock patients
Septic shock Non-septic shock P and yet their NT-proBNP levels did not differ. They also received higher
N = 38 n = 22
norepinephrine doses. In Table 3, the standardized regression coeffi-
Age, years 67 ± 19 69 ± 13 0.82 cients are reported for NT-proBNP levels versus general and cardiac
Male/female sex, no. (%) 20 (53)/18 (47) 15 (68)/7 (32) 0.29
variables. It shows that NT-proBNP levels are affected by age and
Body mass index, kg/m2 25 ± 4 27 ± 4 0.17
APACHE II 27 ± 13 24 ± 11 0.019 PRSW in both septic and non-septic shock, and that creatinine levels
SOFA 10 ± 6 10 ± 4 0.77
Comorbidity
Cardiovascular 11 (29) 16 (73) 0.001
Respiratory 3 (8) 0 0.29 Table 3
Renal 4 (11) 2 (9) 1.0 Standardized regression coefficient for NT-proBNP versus general and cardiac variables
Neurologic 3 (8) 1 (1) 1.0 over 72 h.
Diagnostic group
Sepsis, respiratory 13 (34) - Septic shock P Non-septic shock P
Abdominal 17 (45) - General
Other 8 (21) Age, years 0.41 0.004 0.38 0.028
Non-sepsis, nonsurgical – 7 (32) Gender – 0.99 – b0.001
Cardiac surgery – 6 (27) BMI, kg/m2 0.08 0.48 −0.37 0.019
Vascular surgery – 4 (18) SOFA 0.17 b0.001 0.07 0.68
Trauma – 1 (1) Creatinine, μmol/L 0.38 0.001 0.00 0.49
Other surgery – 4 (18) Norepinephrine, μg/kg/min 0.40 b0.001 0.13 0.16
Associated micro-organisms in sepsis
Gram-positive 6 (16) – Cardiac
Gram-negative 14 (37) – CVP, mm Hg 0.10 0.086 −0.03 0.38
Other 5 (13) – GEDVI, mL/m2 0.25 0.23 0.00 0.17
Bacteremia 12 (31) – Compliance, mm Hg/mL/m2 −0.09 0.016 0.16 0.28
Length of stay ICU, day 17 ± 18 30 ± 24 0.003 LVSWI, cJ/m2 −0.26 0.001 −0.16 0.085
Mortality day 28 17 (45) 5 (28) 0.10 GEF −0.37 b0.001 −0.14 0.38
Mortality ICU 16 (42) 5 (28) 0.17 PRSW, mm Hg −0.43 b0.001 −0.21 0.007
In-hospital mortality 20 (53) 5 (28) 0.031
Abbreviations: BMI, body mass index; SOFA, sequential organ failure assessment; CVP,
Means ± SD or number (percentage), where appropriate. Abbreviations: APACHE, acute central venous pressure; GEDVI, global end-diastolic volume index; LVSWI, left ventricular
physiology and chronic health evaluation; SOFA, sequential organ failure. stroke work index; GEF, global ejection fraction; PRSW, preload-recruitable stroke work.
 A.B.J. Groeneveld, R.J. Trof / IJC Metabolic & Endocrine 10 (2016) 30–35 33

Table 4 3.4. Mortality

Multivariate generalized estimating equations to predict NT-proBNP over 72 h.

Standardized regression P Baseline NT-proBNP levels were higher in non-survivors than in

coefficient survivors at day 28 (P = 0.032). Over the 72 h study period, NT-
Age, year 0.17 0.07 proBNP levels were higher in 28 day non-survivors than survivors
Gender, m/f – 0.08 (P = 0.038), independent of septic/non-septic shock (P = 0.65) and
BMI, kg/m2 −0.10 0.23 APACHE II score (P = 0.29). ICU and hospital non-survivors also had
Creatinine, μmol/L 0.19 0.006
higher baseline NT-proBNP levels than survivors (P = 0.06).
Septic/non-septic shock – 0.62
Norepinephrine, μg/kg/min 0.11 0.16
Compliance, mL/mm Hg/m2 0.03 0.52 4. Discussion
Preload-recruitable stroke work, mm Hg −0.27 0.002

Abbreviations: BMI, body mass index. The current study suggests that NT-proBNP levels are elevated in
septic and non-septic shock and associated, in time, with systolic rather
than diastolic left ventricular dysfunction, assessed by TPTD. Elevations
and norepinephrine doses and thus SOFA scores were associated with prognosticate a poor outcome.
the prohormone levels particularly in septic shock. In multivariate GEE We used the GEDVI as a parameter to normalize stroke volume and
(Table 4), NT-proBNP levels in the 72 h course of the study primarily work, to yield a measure of GEF and PRSW, respectively, as done before
related to PRSW, irrespective of shock origin and independently of [39]. Indeed, TPTD-derived function indices correlate with left ventricu-
other variables that could affect the prohormone level (Table 3). Fig. 1 lar function indices at echocardiography, in the absence of severe right
shows the inverse linear correlation between the variables for pooled heart dysfunction [43,44]. An argument in favor of the latter is the ab-
data, which were log-transformed for the sake of clarity. LVSWI sence of a relation between CVP and NT-proBNP, since right ventricular
(r = − 0.37, P b 0.001) and GEF also inversely correlated to NT- overloading would increase both, even though NT-proBNP may be more
proBNP (r = −0.38, P b 0.001). GEF and PRSW correlated at r = 0.90, sensitive to left than right ventricular dysfunction [6]. Moreover, the
P b 0.001 (n = 175). myocardial depression of sepsis affects right and left ventricles similarly
[45], so that left ventricular dilatation is expected to contribute to
3.3. Prediction GEDVI. Conversely, our results obtained longitudinally and in a relative-
ly large cohort of septic patients, confirm the earlier finding of a relation
We took a GEF b 20% and a corresponding (from linear regression) between hormone levels and (TPTD-derived) indices of systolic rather
PRSW b 15 mm Hg as evidence for systolic cardiac dysfunction. The than diastolic left ventricular dysfunction, irrespective of underlying
AUROC of NT-proBNP for GEF b 20% was 0.69 (P b 0.001), with 94% spec- condition (39). This is in line with some literature [4,5,11,15,20,23,24,
ificity, 36% sensitivity, 85% positive predictive value and 53% negative 32,35,36], but not with others suggesting a primarily inflammatory ori-
predictive value at a cutoff of 6248 pg/mL. The AUROC of NT-proBNP gin of increased plasma NT-proBNP levels in sepsis [2,12,31,33,34]. The
for PRSW b15 mm Hg was 0.72 (P b 0.001) with specificity of 55%, inverse relation with LVSWI agrees with the literature [5]. Our data also
sensitivity of 82%, positive predictive value of 68% of and negative pre- agree with studies suggesting an inverse relationship between NT-
dictive value of 72% at a cutoff of 1480 pg/mL. proBNP elevations and fractional area contraction or ejection fraction

Fig. 1. Relation between (logarithmically converted) NT-proBNP levels and preload-recruitable stroke work (PRSW, mm Hg), for pooled data (n = 145) in the 72 h period: r = −0.45,
P b 0.001.
34 A.B.J. Groeneveld, R.J. Trof / IJC Metabolic & Endocrine 10 (2016) 30–35

(by echocardiography) in sepsis [4,11,14,24,36]. In our study, the in- [6] L. Bal, S. Thierry, E. Brocas, A. van de Louw, J. Pottecher, S. Hours, M.H. Moreau, D.P.
Perrin Gachadoat, A. Tenaillon, B-type natriuretic peptide (BNP) and N-terminal-
verse relation between NT-proBNP with PRSW was even somewhat proBNP for heart failure diagnosis in shock or acute respiratory distress, Acta
stronger than that with GEF, suggesting that NT-proBNP is sensitive to Anaesthesiol. Scand. 50 (2006) 340–347.
changes in systolic cardiac function in sepsis, irrespective of loading [7] R. Rana, N.E. Vlahakis, C.E. Daniels, A.S. Jaffe, G.G. Klee, R.D. Hubmayr, O. Gajic, B-type
natriuretic peptide in the assessment of acute lung injury and cardiogenic edema,
conditions (40,41). Crit. Care Med. 34 (2006) 1941–1946.
The current study, in contrast with our previous one [39], suggests [8] A. Lefebvre, S. Kural-Menasché, M. Darmon, G. Thiéry, J.P. Feugeas, B. Schlemmer, E.
no difference in left ventricular function between septic and non- Azoulay, Use of N-terminal pro-brain natriuretic peptide to detect cardiac origin in
critically ill cancer patients with acute respiratory failure, Intensive Care Med. 34
septic shock patients. However, the myocardial depression of sepsis
(2008) 833–839.
may have somewhat been obscured in this comparison because of [9] R.C. Schutt, C. Cevik, M.P. Phy, Plasma N-terminal prohormone brain natriuretic
more cardiac premorbidity and probably more frequent preexistent peptide as a marker for postoperative cardiac events in high-risk patients undergo-
ing noncardiac surgery, Am. J. Cardiol. 104 (2009) 137–140.
left ventricular dysfunction in the non-septic shock group. In any case,
[10] L. Zapata, P. Vera, A. Roglan, I. Gich, J. Ordonez-Llanos, A.J. Betbesé, B-type natriuretic
the higher NT-proBNP levels associated with septic shock than non- peptides for prediction and diagnosis of weaning failure from cardiac origin, Inten-
septic shock reported in the literature are not confirmed in the current sive Care Med. 37 (2011) 477–485.
study [31]. Postoperative elevations of NT-proBNP have been docu- [11] J. Charpentier, C.-E. Luyt, Y. Fulla, C. Vinsonneau, A. Cariou, S. Grabar, J.-F. Dhainaut,
J.-P. Mira, J.-D. Chiche, Brain natriuretic peptide: a marker of myocardial dysfunction
mented after cardiac, vascular and other types of surgery and to prog- and prognosis during severe sepsis, Crit. Care Med. 32 (2004) 660–665.
nosticate postoperative (cardiac) complications, even though causes of [12] F. Kerbaul, R. Giorgi, C. Oddoze, F. Collart, C. Guidon, P.J. Lejeune, J. Villacorta, F.
the postoperative increases often remained obscure [9,12,16,17,25,26, Gouin, High concentrations of N-BNP are related to non-infectious severe SIRS
associated with cardiovascular dysfunction after off-pump coronary artery surgery,
28]. Our study suggests left ventricular dysfunction as a contributing Br. J. Anaesth. 93 (2004) 639–644.
factor when these postoperative patients manifest shock and need for [13] R.H. Tung, C. Garcia, A.M. Morss, R.M. Pino, M.A. Fifer, B.T. Thompson, K. Lewandrowski,
postoperative resuscitation. This agrees with some studies [17], where- E. Lee-Lewandrowski, J.L. Januzzi, Utility of B-type natriuretic peptide for the evaluation
of intensive care unit shock, Crit. Care Med. 32 (2004) 1643–1647.
as others stressed the relationship with postoperative inflammatory [14] M. Brueckmann, G. Huhle, S. Lang, K.K. Haase, T. Bertsch, C. Weiss, J.J. Kaden, C.
changes and prognosis, at least after cardiac surgery [12]. Finally, in Putensen, M. Borggrefe, U. Hoffmann, Prognostic value of plasma N-terminal pro-
our cohort, NT-proBNP levels were of independent prognostic signifi- brain natriuretic peptide in patients with severe sepsis, Circulation 112 (2005)
527–534.
cance, regardless of underlying disease and its severity, as reported
[15] U. Hoffmann, M. Brueckmann, T. Bertsch, M. Wiessner, C. Liebetrau, S. Lang, K.K. Haase,
before [13–15,20,22–24,27,32,36]. M. Borggrefe, G. Huhle, Increased plasma levels of NT-proANP and NT-proBNP as
Limitations of the study include the imperfect predictive values of markers of cardiac dysfunction in septic patients, Clin. Lab. 51 (2005) 373–379.
[16] A. Mekontso-Dessap, L. Tual, M. Kirsch, G. d'Honneur, D. Loisance, L. Brochard, J.-L.
increased NT-proBNP levels for low GEF and PRSW, possibly because
Teboul, B-type natriuretic peptide to assess haemodynamic status after cardiac sur-
of concomitant confounders or measurement errors (including that of gery, Br. J. Anaesth. 97 (2006) 777–782.
TPTD). This should be weighted, in clinical practice, against the relative [17] S. Provenchère, C. Berroeta, C. Reynaud, G. Baron, I. Poirier, J.-M. Desmonts, B.
ease of the measurements and help in clinical decision making when, Iung, M. Dehoux, I. Philip, J. Bénessiano, Plasma brain natriuretic peptide and
cardiac troponin I concentrations after adult cardiac surgery: association with
for instance, invasive hemodynamic monitoring is not yet applied. We postoperative cardiac dysfunction and 1-year mortality, Crit. Care Med. 34
did not evaluate cardiac function in the other arm of the randomized (2006) 995–1000.
study, monitored with help of a pulmonary artery catheter, since this [18] A.S. McLean, S.J. Huang, S. Hyams, G. Poh, M. Nalos, M. Balik, B. Tang, I. Seppelt, Prog-
nostic values of B-type natriuretic peptide in severe sepsis and septic shock, Crit.
does not allow calculation of GEF and PRSW. Care Med. 35 (2007) 1019–1026.
In conclusion, our current, serial data suggest that NT-proBNP eleva- [19] B. Meyer, M. Huelsmann, P. Wexberg, G. Delle Karth, R. Berger, D. Moertl, T.
tions in critically patients with circulatory shock point to systolic left Szekeres, R. Pacher, G. Heinz, N-terminal pro-B-type natriuretic peptide is an inde-
pendent predictor of outcome in an unselected cohort of critically ill patients, Crit.
ventricular dysfunction, irrespective of the type of shock, and to a Care Med. 35 (2007) 2268–2273.
poor outcome thereof. They may help to recognize and follow cardiac [20] D. Mokart, A. Sannini, J.-P. Brun, D. Blaise, J.-L. Blache, C. Faucher, N-terminal pro-
dysfunction in shock and help its management, when invasive hemody- brain natriuretic peptide as an early prognostic factor in cancer patients developing
septic shock, Crit. Care 11 (2007) R37.
namic monitoring is not yet instituted. [21] E.P. Rivers, J. McCorrd, R. Otero, G. Jacobsen, M. Loomba, Clinical utility of B-type na-
triuretic peptide in early severe sepsis and septic shock, J. Intensive Care Med. 22
(2007) 363–373.
Conflicts of interest [22] M. Varpula, K. Pulkki, S. Karlsson, E. Ruokonen, V. Pettilä, for the FINNSEPSIS study
group, Predictive value of N-terminal pro-brain natriuretic peptide in severe sepsis
None. and septic shock, Crit. Care Med. 35 (2007) 1277–1283.
[23] E. Kandil, J. Burack, S. Sawas, H. Bibawy, A. Schwartzman, M.E. Zenilman, M.H. Bluth,
B-type natriuretic peptide. A biomarker for the diagnosis and risk stratification of
patients with septic shock, Arch. Surg. 143 (2008) 242–246.
Acknowledgments [24] F. Post, L.S. Weilemann, C.-M. Messow, C. Sinning, T. Münzel, B-type natriuretic pep-
tide as a marker for sepsis-induced myocardial depression in intensive care patients,
We thank the staff of the intensive care units for their help in Crit. Care Med. 36 (2008) 3030–3037.
[25] G. Crescenzi, G. Landoni, E. Bidnami, I. Belloni, C. Biselli, C. Rosica, F. Guarracino, G.
collecting the data. Marino, A. Zangrillo, N-terminal B-natriuretic peptide after coronary artery bypass
graft surgery, J. Cardiothorac. Vasc. Anesth. 23 (2009) 147–150.
[26] L. Vetrugno, M.G. Costa, L. Pompei, P. Chiarandini, D. Drigo, F. Bassi, N. Gonano, R.
References Muzzi, G. Della Rocca, Prognostic power of pre- and postoperative B-type natriuretic
peptide levels in patients undergoing abdominal aortic surgery, J. Cardiothorac.
[1] A.S. McLean, S.J. Huang, M. Nalos, B. Tang, D.E. Stewart, The confounding effects of Vasc. Anesth. 26 (2012) 637–642.
age, gender, serum creatinine, and electrolyte concentrations on plasma B-type [27] F. Wang, Y. Wu, L. Tang, W. Zhu, F. Chen, T. Xu, L. Bo, J. Li, X. Deng, Brain natriuretic
natriuretic peptide concentrations in critically ill patients, Crit. Care Med. 31 peptide for prediction of mortality in patients with sepsis: a systematic review and
(2003) 2611–2618. meta-analysis, Crit. Care 16 (2012) R74.
[2] M. Maeder, T. Fehr, H. Rickli, P. Amman, Sepsis-associated myocardial dysfunction: [28] R.N. Rodseth, B.M. Biccard, Y.L. Manach, D. Sessler, G.A.L. Buse, L. Thabane, R.C.
diagnostic and prognostic impact of cardiac troponins and natriuretic peptides, Schutt, D. Bolligerr, L. Caginig, D. Cardinale, C.P.W. Chong, R. Chu, M. Cnotliwy,
Chest 129 (2006) 1349–1366. S.D. Somma, R. Fahrner, W.K. Lim, E. Mahla, R. Manikandan, F. Puma, W.B. Pyun,
[3] J.E. Levitt, A.G. Vinayak, B.K. Gehlbach, A. Pohlman, W. van Cleve, J.B. Hall, J.P. Kress, M. Radović, S. Rajagopalan, S. Suttie, T. Vanniyasingram, W.J.v. Gaal, M. Waliszek,
Diagnostic utility of B-type natriuretic peptide in critically ill patients with P.J. Deveraux, The prognostic value of pre-operative and post-operative B-type na-
pulmonary edema: a prospective cohort study, Crit. Care 12 (2008) R3. triuretic peptides in patients undergoing non-cardiac surgery, J. Am. Coll. Cardiol.
[4] P.R. Forfia, S.P. Watkins, J.E. Rame, Relationship between B-type natriuretic peptides 63 (2014) 170–180.
and pulmonary capillary wedge pressure in the intensive care unit, J. Am. Coll. [29] S. Watanabe, J. Shite, H. Takaoka, T. Shinke, Y. Imuro, T. Ozawa, H. Otake, D.
Cardiol. 45 (2005) 1667–1671. Matsumoto, D. Ogasawara, O.L. Paredes, M. Yokoyama, Myocardial stiffness is an
[5] D. Jefic, J.W. Lee, D. Jefic, R.T. Savoy-Moore, H.S. Rosman, Utility of B-type natriuretic important determinant of the plasma brain natriuretic peptide concentration in
peptide and N-terminal pro B-type natriuretic peptide in evaluation of respiratory patients with both diastolic and systolic heart failure, Eur. Heart J. 27 (2006)
failure in critically ill patients, Chest 128 (2005) 288–295. 832–838.
 A.B.J. Groeneveld, R.J. Trof / IJC Metabolic & Endocrine 10 (2016) 30–35 35

[30] J.L. Januzzi, A. Morss, R. Tung, R. Pino, M.A. Fifer, B.T. Thompson, E. Lee- [38] R. Witthaut, C. Busch, P. Fraunberger, A. Walli, D. Seidel, G. Pilz, R. Stuttmann, N.
Lewandrowski, Natriuretic peptide testing for the evaluation of critically ill patients Speichermann, L. Verner, K. Werdan, Plasma atrial natriuretic peptide and brain
with shock in the intensive care unit: a prospective cohort study, Crit. Care 10 natriuretic peptide are increased in septic shock: impact of interleukin-6 and
(2006) R37. sepsis-associated left ventricular dysfunction, Intensive Care Med. 29 (2003)
[31] B. Wolff, D. Haase, P. Lazarus, K. Machill, B. Graf, H.-G. Lestin, D. Werner, Severe sep- 1696–1702.
tic inflammation as a strong stimulus of myocardial NT-pro brain natriuretic peptide [39] K.J. Hartemink, J.W. Twisk, A.B. Groeneveld, High circulating N-terminal pro-B-type
release, Int. J. Cardiol. 122 (2007) 131–136. natriuretic peptide is associated with greater systolic cardiac dysfunction and
[32] I. Coquet, M. Darmon, J.M. Doise, M. Degrès, B. Blettery, B. Schlemmer, P. Gambert, nonresponsiveness to fluids in septic vs nonseptic critically ill patients, J. Crit. Care
J.P. Quenot, Performance of N-terminal-pro-B-type natriuretic peptide in critically 26 (2011) 108 (e1–8).
ill patients: a prospective observational cohort study, Crit. Care 12 (2008) R137. [40] M. Takeuchi, M. Odake, H. Takaoka, Y. Hayashi, M. Yokoyama, Comparison between
[33] R. Shor, Y. Rozenman, A. Bolshinsky, D. Harpaz, Y. Tilis, Z. Matas, A. Fux, M. Boaz, A. preload recruitable stroke work and the end-systolic pressure–volume relationship
Halabe, BNP in septic patients without myocardial dysfunction, Eur. J. Intern. Med. in man, Eur. Heart J. 13E (1992) 80–84.
17 (2006) 536–540. [41] C.S. Hayward, W.V. Kalnins, R.P. Kelly, Gender-related differences in left ventricular
[34] A. Rudiger, M. Fischler, P. Harpes, S. Gasser, T. Hornemann, A. von Eckardstein, M. chamber function, Cardiovasc. Res. 49 (2001) 340–350.
Maggiorini, In critically ill patients, B-type natriuretic peptide (BNP) and N- [42] R.J. Trof, A. Beishuizen, A.D. Cornet, R.J. de Wit, A.R. Girbes, A.B. Groeneveld, Volume-
terminal pro-BNP levels correlate with C-reactive protein values and leukocyte limited versus pressure-limited hemodynamic management in septic and nonseptic
counts, Int. J. Cardiol. 126 (2008) 28–31. shock, Crit. Care Med. 40 (2012) 1177–1185.
[35] K.L. Turner, L.J. Moore, S.R. Todd, J.F. Sucher, S.A. Jones, B.A. McKinley, A. Valdivia, [43] A. Combes, J.B. Berneau, C.E. Luyt, J.L. Trouillet, Estimation of left ventricular systolic
R.M. Sailors, F.A. Moore, Identification of cardiac dysfunction in sepsis with B-type function by single transpulmonary thermodilution, Intensive Care Med. 30 (2004)
natriuretic peptide, J. Am. Coll. Surg. 213 (2011) 139–146. 1377–1383.
[36] G. Landesberg, P.D. Levin, D. Gilon, S. Goodman, M. Georgieva, C. Weissman, A.S. [44] J. Jabot, X. Monnet, L. Bouchra, D. Chemla, C. Richard, J.-L. Teboul, Cardiac function
Jaffe, C.L. Sprung, V. Barak, Myocardial dysfunction in severe sepsis and septic index provided by transpulmonary thermodilution behaves as an indicator of left
shock — no correlation with inflammatory cytokines in real-life clinical setting, ventricular systolic function, Crit. Care Med. 37 (2009) 2913–2918.
Chest 148 (2015) 93–102. [45] M.M. Parker, K.E. McCarthy, F.P. Ognibene, J.E. Parrillo, Right ventricular dysfunction
[37] G. Vila, M. Resl, D. Stelzeneder, J. Struck, C. Maier, M. Riedl, M. Hülsmann, R. Pacher, and dilatation, similar to left ventricular changes, characterize the cardiac
A. Luger, M. Clodi, Plasma NT-proBNP increases in response to LPS administration in depression of septic shock in humans, Chest 97 (1990) 126–131.
healthy men, J. Appl. Physiol. 105 (2008) 1741–1745.