1812 www.ccmjournal.

org August 2014 Volume 42 Number 8

Objective: The frequency, risk factors, and mortality rates of seri-
ous adverse events associated with the use of vasopressin and
norepinephrine are not clear. The objectives of this study were to
determine frequency, risk factors (including candidate gene poly-
morphisms), and outcomes of serious adverse events in septic
shock patients.
Design: Retrospective cohort study using multicenter discovery
and single-center validation cohorts.
Setting: ICUs at academic teaching centers.
Patients: Five hundred ninety-seven patients with septic shock in
discovery (Vasopressin and Septic Shock trial) and 533 patients
in validation (St. Pauls Hospital) cohorts.
Intervention: Vasopressin and norepinephrine for septic shock.
Measurements and Main Results: The primary outcome variable
was 90-day mortality rates of patients with and without serious
adverse events. Secondary outcome variables were the association
between vasopressor genotype pathway polymorphisms, plasma
vasopressin levels, and serious adverse events. Plasma vasopres-
sin concentrations were measured at baseline, 6 hours, 24 hours,
72 hours, and 7 days after vasopressor infusion. Patients with sep-
tic shock were genotyped for 268 vasopressor pathway tag sin-
gle-nucleotide polymorphisms. Serious adverse events occurred
in 10.5% and 9.7% of patients in Vasopressin and Septic Shock
trial and St. Pauls Hospital cohorts, respectively. Patients who
had serious adverse events had higher mortality (p < 0.01) than
patients without serious adverse events (adjusted for age, serum
lactate, Acute Physiology and Chronic Health Evaluation II, and
maximum dose of norepinephrine day 1) (hazard ratio, 2.97;
95% CI, 2.204.00; p < 0.001 and hazard ratio, 1.89; 95% CI,
1.262.85; p = 0.002 in Vasopressin and Septic Shock trial and
St. Pauls Hospital, respectively). There was no difference in the
area under the plasma vasopressin concentration curve between
patients with and without serious adverse events (p = 0.1). The
AA genotype of rs28418396 single-nucleotide polymorphism
(near the arginine vasopressin receptor 1b gene) was signifcantly
associated with serious adverse events in discovery and valida-
tion cohorts (p = 0.001 and p = 0.04, respectively).
Conclusion: Serious adverse events associated with vasopres-
sin and norepinephrine in patients who have septic shock are
associated with increased mortality and morbidity. AA genotype
of rs28418396 single-nucleotide polymorphism near the arginine
vasopressin receptor 1b gene is associated with serious adverse
events. The mechanism of this association requires investigation.
(Crit Care Med 2014; 42:18121820)
Key Words: arginine vasopressin receptor 1b; norepinephrine;
septic shock; serious adverse events; vasopressin
I
n the latest surviving sepsis guidelines, norepinephrine
(NE) is recommended as rst vasopressor and vasopres-
sin can be added to NE with intent of either raising mean
arterial pressure (MAP) or decreasing NE dosage (1). However,
these commonly used vasopressors are associated with serious
adverse events (SAEs) (2). Although the hemodynamic status
may be improved by vasopressors, several studies of vasopres-
sin and NE infusions found the prevalence of adverse drug
Copyright 2014 by the Society of Critical Care Medicine and Lippincott
Williams & Wilkins
DOI: 10.1097/CCM.0000000000000333
*See also p. 1944.
1
Centre for Heart Lung Innovation, St. Pauls Hospital, Vancouver, BC,
Canada.
2
Division of Critical Care Medicine, St. Pauls Hospital, University of British
Columbia, Vancouver, BC, Canada.
Registration: ISRCTN94845869.
Supplemental digital content is available for this article. Direct URL cita-
tions appear in the printed text and are provided in the HTML and PDF
versions of this article on the journals website (http://journals.lww.com/
ccmjournal).
Supported, in part, by VASST, Canadian Institutes of Health Research
(grant number: MCT 44152).
Dr. Boyd is a recipient of a Providence Health Care Research Scholar-
ship. Dr. Walley is employed by the University of British Columbia and has
stock options with Sirius Genomics. His institution received grant support
from the University of British Columbia and Canadian Institutes of Health
Research. Dr. Russell served as board member for Sirius Genomics; con-
sulted for Ferring Pharmaceutical, Grifols, Medimmune, and Astra Zeneca;
lectured for Pfzer; has a patent with the University of British Columbia and
Sirius Genomics; and has stock options with Sirius Genomics. His institu-
tion received grant support from Ferring Pharmaceutical, Astra Zeneca,
Sirius Genomics, and Eli Lilly. Dr. Anantasit has disclosed that he does not
have any potential conficts of interest.
For information regarding this article, E-mail: jim.russell@hli.ubc.ca
Serious Adverse Events Associated With Vasopressin
and Norepinephrine Infusion in Septic Shock*
Nattachai Anantasit, MD
1
; John H. Boyd, MD, FRCP(C)
1,2
; Keith R. Walley, MD, FRCP(C)
1,2
;
James A. Russell, MD, FRCP(C)
1,2
Clinical Investigations
Critical Care Medicine www.ccmjournal.org 1813
events ranged from 10% to 72% (36). SAEs of vasopressors
include arrhythmias, chest pain, myocardial infarction, cardiac
arrest, decreased cardiac output, stroke, limb ischemia, mesen-
teric ischemia, and water intoxication (711). SAEs occurred in
about 10% of patients in a randomized controlled trial of vaso-
pressin versus NE (Vasopressin and Septic Shock trial, VASST),
with no difference in frequency between treatment groups (5).
The risk of these SAEs in VASST has not yet been investigated.
SAEs can lead to death, but it is not well understood whether
and to what extent patients who have SAEs associated with
vasopressin and NE have an increased risk of death.
In addition, SAEs associated with vasopressin or NE infu-
sion in patients with septic shock can be difcult to predict.
Clinical markers, vasopressin levels, and pharmacogenomic
markers could identify patients who have an increased risk
of SAEs and the clinician can then avoid or discontinue that
drug. Genetic single-nucleotide polymorphisms (SNPs)
are important identiable risk factors in patients who may
develop SAEs (12).
Limited lists of candidate genes in the vasopressin and NE
pathways modulate their actions and so could be candidates
for evaluation as SNPs predictive of SAEs associated with vaso-
pressin and NE. Vasopressin stimulates a family of receptors
including arginine vasopressin receptor (AVPR)1a (V1 recep-
tor), AVPR1b (V3 receptor), AVPR2 (V2 receptor), oxytocin
receptor, and purinergic receptors (13). Vasopressinase (leucyl/
cystinyl aminopeptidase, LNPEP) is associated with mortality
and increased vasopressin clearance in septic shock patients
(14). NE stimulates adrenergic receptors. The -adrenergic
receptor gene (ADR) plays an important role in cardiovascular
disease and SNPs of ADR have been associated with mortality
of septic shock (15). Furthermore, the angiotensin II signaling
pathway gene polymorphisms are associated with outcomes in
cardiovascular disorders (16). An angiotensin II type 1 recep-
torassociated protein (AGTRAP) polymorphism is associated
with mortality in patients with septic shock (17).
Our overarching goal was to determine the mortality rate of
SAEs associated with vasopressin and NE infusion in patients
who have septic shock. We also sought to determine whether
commonly available clinical characteristics are associated with
an increased risk of developing SAEs. Finally, we hypothesized
that SNPs of candidate genes of the vasopressin and NE path-
ways and/or plasma arginine vasopressin (AVP) levels could
predict SAEs in patients who had septic shock who received
vasopressin versus NE infusions.
MATERIALS AND METHODS
Patients
VASST Cohort. We used the VASST cohort as the discov-
ery cohort because VASST was a multicenter, randomized,
double-blind, controlled trial evaluating the safety and efcacy
of vasopressin 0.010.03 U/min versus NE 515 g/min in a
total of 778 patients with septic shock. Entry criteria required
that eligible patients were older than 16 years, had two of four
criteria for systemic inammatory response syndrome (SIRS),
had a new proven or suspected infection, had new dysfunction
of at least one new organ dysfunction, and had hypotension
despite adequate uid resuscitation on a minimum of 5 g/
min of NE at baseline (5). During the initiation and titration
of the study drug, the bedside nurse also titrated open-label
vasopressors to maintain a constant target MAP. An initial tar-
get MAP of 6575 mm Hg was recommended. Of these, 597
patients had complete patient data, DNA available, and were
included in the current study. The research ethics boards of
all participating institutions approved this trial, and writ-
ten informed consent was obtained from all patients or their
authorized representatives. The research ethics board at the
coordinating center (University of British Columbia) approved
the genetic analysis.
St. Pauls Hospital Cohort. We used the St. Pauls Hospi-
tal (SPH) cohort as a validation cohort because it included
patients in whom vasopressin and NE were used in prac-
tice. However, vasopressin and NE use were not randomized,
blinded, or protocolized. We did a retrospective chart review
of patients who admitted at SPHICU with septic shock. Septic
shock was dened by infection, the presence of two or more
diagnostic criteria for the SIRS, and hypotension despite ade-
quate uid resuscitation or requiring vasopressor (18). Vaso-
pressin was generally used after patients were already started
on NE infusions (as rst vasopressor) usually, when patients
were not responsive and needed increased doses of vasopressin.
There was no difference in use of vasopressin over the 5 years.
Patients admitted to the ICU of SPH in Vancouver, Canada,
between July 2000 and January 2004 were screened (n = 1,626).
Of these, 533 patients had septic shock at admission, were
extensively phenotyped, and had DNA available. Phenotyping
and genotyping were not part of routine practice at SPH. The
institutional review board at SPH and the University of British
Columbia approved the study. This study used discarded blood
and anonymized databases, and so the ethics board ruled that
consent was not necessary.
Denitions
SAEs were dened as any severe, life-threatening event tem-
porally associated with the use of vasopressin or NE. Acute
myocardial ischemia was dened by serum cardiac troponin
T above normal plus electrocardiogram (ECG) ndings sug-
gesting myocardial ischemia as interpreted by clinical cardi-
ologists interpreting ECGs. Tachyarrhythmia was dened as a
new onset of atrial brillation. Bradyarrhythmia was dened
as a new heart rate below 50 beats/ min which require rapid
intervention. Cardiac arrest dened as asystole, ventricular
brillation, or pulseless electrical activity. Digital ischemia was
dened as the occurrence of new areas of mottled or livido
skin in one or more extremities (4). Mesenteric ischemia was
dened by evidence of radiographic signs of ischemic bowel
on abdominal radiograph as interpreted by clinical radiolo-
gists. Cerebrovascular accident was dened as new neurologic
decit and acute cerebral infarct or hemorrhage on CT brain.
Hyponatremia was dened as a serum sodium level of less than
130 mmol/L.
Anantasit et al
1814 www.ccmjournal.org August 2014 Volume 42 Number 8
We evaluated organ dysfunction by calculating days alive
and free (DAF) of organ dysfunction (19). Organ dysfunction
for each organ system was considered to be present during
each 24-hour period if there was evidence of moderate, severe,
or extreme organ dysfunction according to the Brussels cri-
teria (20). Each day a patient was given 1 if alive and free of
organ dysfunction and zero if dead or had organ dysfunction.
We also similarly assessed DAF ventilation, vasopressors, and
renal support over 28 days. The lowest score possible for each
variable was zero and the highest score possible was 28. A low
score indicates more organ dysfunction, whereas a high score
indicates less organ dysfunction.
Plasma Vasopressin Levels
Blood for measurement of plasma vasopressin levels was col-
lected at baseline, 6 hours, 24 hours, 72 hours, and 7 days in
VASST. Plasma vasopressin concentration was measured
after extraction on reversed phase column by double anti-
body immunoassay (5). This subgroup of 88 patients who
had plasma vasopressin concentrations measured (n = 45 in
vasopressin group and n = 43 in NE group) had similar clini-
cal characteristics to the original VASST cohort (subgroup vs
original VASST cohort, median age [yr]: 62 vs 63; male [%]:
60.2 vs 60.9; Acute Physiology and Chronic Health Evalua-
tion II [APACHE II]: 26 vs 27; serum lactate [mmol/L]: 2.1
vs 2.2). We calculated the area under the plasma vasopressin
concentration-time curve (AUC of the vasopressin levels) of
each patient for later group comparisons.
Selection of SNPs of Vasopressin and NE Pathway
and Genotyping
DNA was extracted from the buffy coat of discarded blood
samples using a QIAamp DNA maxi kit (Qiagen, Mississauga,
ON, Canada) and genotyped using the Illumina Golden Gate
assay (Illumina, San Diego, CA). We used genotyping data
from Seattle SNPs (http://pga.gs.washington.edu/) and Hap-
Map (http://www.hapmap.org/) with an r
2
threshold of 0.65
for SNPs with a minor allele frequency more than 5% to iden-
tify tag SNPs. We identied four tag SNPs (AGTRAP) and 16
tag SNPs (ADR) in NE pathway and three tag SNPs (AVPR1a),
ve tag SNPs (AVPR1b), 237 tag SNPs (LNPEP), and three tag
SNPs (OXTR) in the vasopressin pathway.
Primary and Secondary Outcomes
The primary outcome was 90-day mortality, and we compared
the mortality rates of patients who did or did not have a SAE.
We also evaluated whether baseline clinical characteristics dif-
fered between patients with or without SAEs. Finally, we also
determine whether plasma vasopressin levels and SNPs related
to vasopressin and NE were associated with SAEs.
Statistical Analysis
Statistical analysis was performed using R, version 2.8.1 (http://
www.R-project.org), and SPSS, version 16 (SPSS, Chicago, IL)
statistical software packages. We compared patients who had
developed SAEs with no SAEs and tested for differences in
baseline characteristics and hospital outcomes including DAF
of organ dysfunction by using the Mann-Whitney U statistic
test for continuous data or a chi-square test for categorical data.
Normally distributed continuous variables were described by
mean and SD, whereas nonnormally distributed continuous
variables were expressed by median and interquartile ranges.
Categorical variables were expressed as number of patients and
percentage. We tested for Hardy-Weinberg equilibrium using a
chi-square test primarily as a data quality check. The Armitage
trend test was used for evaluation of the association of SNPs
TABLE 1. Serious Adverse Events in Discovery (Vasopressin and Septic Shock Trial) and
Validation (St. Pauls Hospital) Cohorts
Variable
Vasopressin and Septic Shock
Trial Cohort (n = 597) (%)
St. Pauls Hospital Cohort
c

(n = 533) (%) p
Myocardial ischemia 12 (2) 14 (2.6) 0.49
Tachyarrhythmias 6 (1) 14 (2.6) 0.039
a
Bradyarrhythmias 7 (1.2) 10 (1.9) 0.332
Cerebrovascular accident 2 (0.3) 7 (1.3) 0.093
Unexpected cardiac arrest 7 (1.2) 3 (0.6) 0.349
Digital ischemia 8 (1.3) 3 (0.6) 0.184
Mesenteric ischemia 14 (2.3) 2 (0.4) 0.005
a
Hyponatremia 2 (0.3) 0 0.501
Others
b
6 (1) 0 0.032
a
Total 63 (10.6) 52 (9.7) 0.658
a
p < 0.05.
b
Other events include acute hepatitis, agranulocytosis, pulmonary embolism, seizures, drug error, and drug extravasation from the central venous catheter.
c
Electrolytes were not reviewed in St. Pauls Hospital cohort.
p values are based on Pearson chi-square test.
Clinical Investigations
Critical Care Medicine www.ccmjournal.org 1815
with presence or absence of SAEs. Kaplan-Meier curves were
constructed for comparison of the estimated probability of
survival in the two outcome groups and were tested for differ-
ence using the log-rank test. We also did an analysis comparing
SAE versus no SAE adjusting for important baseline variables
using Cox regression analysis. The occurrence of SAE dur-
ing hospital stay was considered a time-dependent covariate
in the analysis. We tested the above hypotheses in the discov-
ery cohort and then took forward only those baseline clinical
characteristics and SNPs that were signicant in the discovery
cohort for further evaluation in the validation cohort. In all
comparisons, p value of less than 0.05 was considered statisti-
cally signicant.
RESULTS
The frequency of SAEs was similar in the discovery and vali-
dation cohorts (Table 1). There were 64 SAEs in 63 patients
(10.7% of the VASST cohort) and 53 events in 52 patients
(9.9% of the SPH cohort). Mesenteric ischemia (2.3%) was the
most common SAE in VASST, whereas myocardial ischemia
and tachyarrhythmias (2.6%) were the most common SAEs in
SPH (Table 1).
We found that similar proportions of patients on vaso-
pressin versus NE in the discovery and validation cohorts had
SAEs. In VASST, 33 of 304 patients who received vasopressin
(AVP group) (10.8%) had SAEs and 30 of 293 patients who
received NE (NE group) (10.2%) had SAEs. Similarly, in SPH,
14 of 111 patients in AVP group (12.6%) and 38 of 422 patients
in NE group (9%) had SAEs.
In VASST, patients who had developed SAEs were sig-
nicantly older (69 vs 63 yr, p = 0.045) and had a higher ini-
tial arterial lactate level (2.8 vs 2.2 mmol/L, p = 0.035) than
patients who did not develop SAEs. In addition, patients who
had SAEs received signicantly higher (median) NE dose on
day 1 than patients without SAEs (20 g/min vs 15 g/min,
p = 0.016) (Table 2). In the SPH cohort, patients who had SAEs
TABLE 2. Baseline Characteristics of Patients With or Without Serious Adverse Events
With Septic Shock in the Discovery (Vasopressin and Septic Shock Trial) and Validation
(St. Pauls Hospital) Cohorts
Variable
Vasopressin and Septic Shock
Trial Cohort
St. Pauls Hospital
Cohort
No SAE
(n = 534)
Any SAE
(n = 63) p
No SAE
(n = 481)
Any SAE
(n = 52) p
Age, yr 63 (5072) 69 (5475) 0.045
a
62 (4772) 64 (4873) 0.341
Gender: male, n (%) 320 (59.9) 33 (52.4) 0.249 307 (63.8) 31 (59.6) 0.549
Race: Caucasian, n (%) 449 (84.1) 51 (81.0) 0.524 374 (77.8) 42 (80.8) 0.618
Acute Physiology and Chronic Health Evaluation II 26 (2232) 28 (2334) 0.100 26 (2132) 29 (2434) 0.035
a
Use vasopressor: NE 263 (49.3%) 30 (47.6%) 0.806 384 (79.8%) 38 (73.1%) 0.254
Maximum dose of NE on day 1 (g/min) 15 (926) 20 (1134) 0.016
a
12 (525) 15 (625) 0.593
Preexisting condition, n (%)
Congestive heart failure 44 (8.2) 2 (3.2) 0.154 27 (5.6) 5 (9.6) 0.226
Chronic obstructive lung disease 97 (18.2) 8 (12.7) 0.281 78 (16.2) 10 (19.2) 0.578
Chronic renal failure 62 (11.6) 3 (4.8) 0.099 33 (6.9) 2 (3.8) 0.562
Liver disease 57 (10.7) 11 (17.5) 0.109 50 (10.4) 6 (11.5) 0.798
Chronic corticosteroid use 114 (21.3) 8 (12.7) 0.107 28 (5.8) 3 (5.8) 1
Initial hemodynamic
Mean arterial pressure (mm Hg) 55 (5061) 56 (4863) 0.876 54 (5059) 54 (5060) 0.659
Heart rate (beats/min) 126 (110140) 130 (112142) 0.146 115 (95130) 105 (90130) 0.024
a
Serum lactate (mmol/L) 2.2 (1.44.3) 2.8 (1.95.5) 0.035
a
2.3 (1.44.7) 2.9 (2.05.6) 0.054
rs28418396_AVPR1B
TT/TA 150 (28.1%) 8 (12.7%) 0.009
a
134 (27.9%) 6 (11.5%) 0.011
a
AA 384 (71.9%) 55 (87.3%) 347 (72.1%) 46 (88.5%)
SAE = serious adverse event, NE = norepinephrine.
a
p < 0.05.
Data are median (interquartile range) for continuous variables. p values were calculated with the use of Pearson chi-square test and Mann-Whitney U test.
Anantasit et al
1816 www.ccmjournal.org August 2014 Volume 42 Number 8
had higher APACHE II than patients who did not have SAEs
(29 vs 26, p = 0.035) and had lower heart rates (105 vs 115,
p = 0.024) (Table 2). There were no other signicant differ-
ences between groups.
Regarding timing of SAEs (Supplemental Table 1,
Supplemental Digital Content 1, http://links.lww.com/CCM/
A913), the mean time of SAEs after onset of vasopressor
infusion was similar in the vasopressin and NE groups and
ranged from 0 to 26 days. Time to SAE in the VASST and SPH
cohorts is shown in Supplemental Figure 1 (Supplemental
Digital Content 2, http://links.lww.com/CCM/A914). Nearly
all SAEs occurred within 10 days of onset of shock treat-
ment. There was no difference in time to SAE according
to treatment group (vasopressin vs NE) in the VASST or
SPH cohorts (Supplemental Fig. 2, Supplemental Digital
Content 3, http://links.lww.com/CCM/A915; Supplemental
Fig. 3, Supplemental Digital Content 4, http://links.lww.
com/CCM/A916).
Overall, 90-day mortality rates were 44.5% and 54.4% in
VASST and SPH cohorts, respectively. Patients who had SAEs
had higher 90-day (and 28-day) mortality rates (Fig. 1) and
had greater organ dysfunction (cardiovascular, vasopressor use,
ventilation, renal replacement, neurologic, hematologic, and
hepatic) (Table 3). Cox regression analysis to adjust for clinical
variables signicantly associated with SAEs (age, serum lactate,
APACHE II, and maximum dose of NE on day 1) conrmed that
patients who had serious events had higher mortality at 90 days
(hazard ration [HR], 2.97; 95% CI, 2.204.00; p < 0.001 and HR,
1.89; 95% CI, 1.262.85; p = 0.002 in the whole VASST cohort
[n = 779] and SPH, respectively) than patients without SAEs
even after adjusting for these prognostic factors. Cox regression
analysis to adjust for clinical variables signicantly associated
with SAEs (age, serum lactate,
APACHE II, maximum dose of
NE on day 1, and AA genotype
of rs 28418396) conrmed that
patients who had serious events
had higher mortality at 90 days
(HR, 1.654; 95% CI, 1.152.39;
p = 0.007 and HR, 2.08; 95%
CI, 1.373.14; p < 0.001 in
VASST and SPH, respectively)
than patient without SAEs.
Furthermore, Cox regression
analysis to adjust for clinical
variables signicantly associated
with SAEs (age, serum lactate,
APACHE II, maximum dose
of NE on day 1, and any SAE)
conrmed that patients who
had AA genotype of rs28418396
had higher mortality at 90 days
(HR, 1.41; 95% CI, 1.061.88;
p = 0.018 and HR, 1.53; 95% CI,
1.132.07; p = 0.006 in VASST
and SPH, respectively) than
patient who had TA/TT geno-
type of rs28418396.
Comparing the vasopres-
sin with NE groups in VASST
and then in SPH, patients with
SAEs had higher mortality and
fewer DAF of organ dysfunc-
tion in most comparisons except
for the vasopressin group in
SPH (Supplemental Table 2,
Supplemental Digital Content
5, http://links.lww.com/CCM/
A917).
The area under the curve of
plasma vasopressin concentra-
tion was not different between
Figure 1. Kaplan-Meier survival curves of septic shock patients who had developed any serious adverse events
(SAEs) compared with septic shock patients who had not developed SAEs in both discovery (Vasopressin and
Septic Shock trial, VASST) and validation (St. Pauls Hospital, SPH) cohorts.
Clinical Investigations
Critical Care Medicine www.ccmjournal.org 1817
patients with and without SAEs (Fig. 2). As expected, the AUC
of AVP level showed that patients in NE group had much lower
AUC than those in the vasopressin group (Fig. 2).
We successfully genotyped 268 tag SNPs in vasopressor path-
way genes in VASST cohort. All SNPs were in Hardy-Weinberg
equilibrium. In VASST, nine SNPs were signicantly associated
with SAEs (Table 4). Of these, only rs28418396 was also signi-
cant in the validation (SPH) cohort. We found that patients who
had AA genotype of rs28418396 in NE group were statistically
signicant increased risk of SAEs in both cohorts (p = 0.001 in
VASST discovery cohort, which is signicant after Bonferroni
correction; p = 0.04 in SPH validation cohort, single SNP vali-
dation so not corrected in validation cohort) (Fig. 3).
An adjusted analysis (multiple logistic regression) in the
discovery cohort (VASST) using age, APACHE II, arterial lac-
tate, and maximum dose of NE on day 1 as covariates showed
that patients who had the AA genotype of rs28418396 SNP had
signicantly more SAEs than patients who had TA/TT geno-
types (AA vs TA/TT, VASST: adjusted odds ratio [OR], 2.27;
95% CI, 1.034.98; p = 0.041). Logistic regression analysis for
SAEs in SPH validated that patients who had AA genotype of
rs28418396 had greater risk of SAEs than patients who had TT/
TA genotypes of rs28418396 (AA vs TT/TA, adjusted OR, 2.99;
95% CI, 1.247.21; p = 0.014 after adjusting for age, serum
lactate, and APACHE II). Patients who had AA genotype of
rs28418396 in vasopressin group had signicant lower AUC of
AVP level over the rst 7 days than patient who had TA/TT
genotype (p = 0.022).
DISCUSSION
In this study, we found that 10% of patients with septic shock
who received vasopressin or NE had SAEs in discovery and val-
idation cohorts. Second, we found that patients who had SAEs
had about double the mortality rates of patients who did not
have SAEs. Third, we found that the area under the vasopres-
sin concentration curve was not different between patients on
vasopressin infusion with or without SAEs. Finally, we found
that patients who had the AA genotype of rs28418396 SNP had
signicantly higher rates of SAEs in both discovery and valida-
tion cohorts.
The SAE rates we found are somewhat lower when com-
pared with other studies (6, 9). One reason could be use of dif-
ferent denitions of SAEs in different studies. Different dosing
of vasopressin and NE is not likely the explanation as the doses
of vasopressin and NE of our study were similar (vasopressin
dose was 0.010.03 U/min; mean maximum doses of NE were
26.7 and 24.6 g/min in VASST and SPH cohort, respectively)
to other studies.
TABLE 3. Mortality Rates and Morbidities of Patients With or Without Serious Adverse
Events in the Discovery (Vasopressin and Septic Shock Trial) and Validation (St. Pauls
Hospital) Cohorts
Variable
Vasopressin and Septic Shock Trial Cohort St. Pauls Hospital Cohort
No SAE (n = 534) Any SAE (n = 63) p No SAE (n = 481) Any SAE (n = 52) p
Outcome, n (%)
28-day mortality 168 (31.5%) 35 (55.6%) < 0.001
a
215 (44.7%) 34 (65.4%) 0.005
a
90-day mortality 228 (42.7%) 38 (60.3%) 0.008
a
251 (52.2%) 39 (75%) 0.002
a
Length of ICU stay (d) 10 (516) 9 (319) 0.62 8 (415) 10 (421) 0.611
Days alive and free of organ dysfunction
Cardiovascular 20 (024) 3 (021) < 0.001
a
11 (023) 4 (016) 0.006
a
Vasopressor use
b
20 (024) 3 (022) 0.001
a
18 (125) 7 (020) 0.004
a
Ventilation
c
10.5 (021) 0 (08) < 0.001
a
3 (020) 1 (09) 0.039
a
Renal replacement 26 (628) 16 (128) 0.006
a
15 (228) 7 (123) 0.042
a
Neurologic
d
18 (025) 2 (018) < 0.001
a
22 (427) 9 (225) 0.007
a
Hematologic
e
25 (728) 19 (128) 0.005
a
21 (428) 10 (225) 0.031
a
Hepatic 27 (728) 11 (128) 0.001
a
20 (328) 8 (127) 0.013
a
Any organ failure 0 (010) 0 (02) 0.012
a
0 (00) 0 (00) 0.216
a
SAE = serious adverse events.
a
p < 0.05.
b
Vasopressor use was defned as 5 g or more of dopamine/kg/min or any dose of norepinephrine, epinephrine, or phenylephrine.
c
Ventilation was defned as intubation and positive-pressure ventilation.
d
Neurologic was defned as Glasgow Coma Score < 12, prior to receiving sedation.
e
Hematologic was defned as new onset of coagulation defect or platelet count < 80,000/mm
3
.
Data are median (interquartile range) for continuous data. p values were calculated with Pearson chi-square test and Mann-Whitney U test.
Anantasit et al
1818 www.ccmjournal.org August 2014 Volume 42 Number 8
This study showed that patients with SAEs had higher (about
double) mortality rate than patients without SAEs. Because
this was an association study, it is not clear whether the SAEs
were in some part causal of the increased mortality rates or
whether the presence of SAEs simply identied sicker patients
at increased risk of death. Some clinical markers were associated
with increased risk of SAEs in one or both cohorts. For example,
patients who had developed SAEs had a trend to higher serum
lactate than patients without SAEs. Prior studies found that high
serum lactate is a risk predictor of nonocclusive mesenteric isch-
emia (21). Patients with SAEs also were older and had higher
APACHE II than patients without SAEs. These simple clinical
markers could be used to iden-
tify patients at increased risk of
SAEs while on vasopressin or NE
infusions for septic shock.
We found that the AA geno-
type of rs28418396 SNP was sig-
nicantly associated with SAEs
in discovery (VASST) and vali-
dation (SPH) cohorts. To our
knowledge, this is the rst study
to demonstrate that AA geno-
type of rs28418396 SNP is a risk
factor for SAEs in patients with
septic shock. The rs28418396 is
near the 5 untranslated region
of the AVPR1b gene. AVPR1b
gene sized is 7,200 bases in
chromosome 1q32. AVP is pro-
duced in the hypothalamus,
stored in posterior pituitary,
and released into the circula-
tion (2). There is an impor-
tant interaction between AVP
and the adrenal axis in stress
situations that is linked by the
AVPR1b receptor (22). When
vasopressin binds to the V1b
receptor on corticotrophs in the
anterior pituitary (coded for by
the AVPR1b gene), adrenocor-
ticotropin hormone (ACTH) is
released (23). Septic shock acti-
vates vasopressin release to sup-
raphysiologic levels and then
vasopressin levels decrease rap-
idly (24). Furthermore, animal
studies show that the activation
of central vasopressin may be
mediated by NE through 1-
adrenergic receptors (25). NE
also triggers the pituitary gland
to release ACTH.
The exact mechanism by which
rs28418396 SNP of AVPR1b gene
inuences any SAEs, particularly patients who received NE infu-
sion, is unknown and requires further investigation.
Plasma vasopressin levels were not associated with SAEs in
septic shock patients. In particular for patients on vasopressin
infusion, vasopressin levels were not different between patients
with or without SAEs suggesting that simple pharmacokinetics
and higher vasopressin levels were not associated with SAEs.
Vasopressin secretion is complex and depends on blood pres-
sure, plasma osmolarity, and blood volume (26), as well as
hypercapnia, hypoxia, hyperthermia, pain, nausea, and mor-
phine. Furthermore, many hormones are direct stimulators
of vasopressin, such as acetylcholine, histamine, angiotensin
Figure 2. The area under the plasma vasopressin concentration-time curve (AUC) in septic shock patients in
the discovery (Vasopressin and Septic Shock trial) cohort at 24 hr and 7 days in the norepinephrine (NE) and
arginine vasopressin (AVP) groups without (top) or with regard to the genotype of rs28418396 of AVPR1b.
SAE = serious adverse events.
Clinical Investigations
Critical Care Medicine www.ccmjournal.org 1819
II, prostaglandins, dopamine, and, particularly, the adrenergic
system. At low concentrations, NE increases vasopressin activ-
ity; conversely, at high concentrations, NE inhibits vasopressin
secretion (27). The half-life of vasopressin is approximately 10
minutes. Therefore, vasopressin levels are not good estimates
of its production and release. This study showed that the area
under the plasma concentration-time curve of vasopressin was
not different between patients with or without SAEs in both
vasopressin and NE groups.
The clinical implications of our study are important. We
suggest that the original VASST randomized controlled trial
provides guidance on efcacy of vasopressin versus NE in that
in the stratum of patients who had less severe shock (dened
as NE infusion < 15 g/min), patients in the vasopressin group
had lower mortality than patients in the NE group. Regarding
safety, the original publication and the current study sug-
gest that the risk of SAEs is similar for vasopressin versus NE.
Furthermore, patients must be proactively monitored for SAEs
because they are frequent and
are associated with increased
mortality. Second, we sug-
gest that vasopressor infusions
should be altered (decrease
dose, switch to an alternative
vasopressor, re-evaluate volume
status, and consider administer-
ing volume to allow vasopressor
weaning). Third, if the genomic
marker rs28418396 is validated
further, then in future, patients
could be assessed by measur-
ing their genotype and altering
vasopressor choice according to
genotype.
There are several limitations
in this study. First, the valida-
tion (SPH) cohort was a retro-
spective chart review (which
may underestimate SAEs seen
in practice) and there were no
plasma vasopressin level mea-
surements. Second, this is an
association study so we can-
not conclude to what extent
TABLE 4. Allele Frequency and Associations With Serious Adverse Events of Signicant
Tag Single-Nucleotide Polymorphisms in Vasopressor Pathway Genes in the Discovery
(Vasopressin and Septic Shock Trial) Cohort
Single-Nucleotide Polymorphism_Gene
Major/Minor
Allele
Minor Allele Frequency
(n = 597)
Hardy-Weinberg
Equilibrium p
Serious Adverse
Event p
rs28418396_AVPR1b A/T 0.14 0.538 0.034
rs11121816_AGTRAP G/T 0.40 0.156 0.031
rs11953730_LNPEPregion C/T 0.001 0.984 0.004
rs11958177_LNPEPregion A/G 0.001 0.984 0.004
rs2549779_LNPEPregion A/G 0.48 0.402 0.012
rs2549780_LNPEPregion A/G 0.48 0.807 0.030
rs17531248_LNPEPregion C/G 0.18 0.474 0.031
rs9127_LNPEPregion A/G 0.42 0.542 0.043
rs27302_LNPEPregion A/G 0.38 0.787 0.044
AVPR1b = arginine vasopressin receptor 1b, AGTRAP = angiotensin II type 1 receptorassociated protein, LNPEP = leucyl/cystinyl aminopeptidase.
p value were calculated with the use of chi-square test for Hardy-Weinberg equilibrium and Breslow generalized Wilcoxon test for serious adverse events.
Figure 3. Percentage of AA genotypes of rs28418396 single-nucleotide polymorphism in septic
shock patients with or without serious adverse events (SAE) according to arginine vasopressin (AVP) or
norepinephrine groups in discovery (Vasopressin and Septic Shock trial, VASST) and validation (St. Pauls
Hospital, SPH) cohorts. *p = 0.001 in VASST and p = 0.04 in SPH.
Anantasit et al
1820 www.ccmjournal.org August 2014 Volume 42 Number 8
the development of SAEs while on vasopressin or NE infu-
sion contributed to increased mortality. Third, we could not
directly address vasopressin production and release. We did,
however, assess the relationship between plasma vasopressin
levels and occurrence of SAEs because higher doses and levels
of vasopressin could have caused greater vasoconstriction and
so could have caused more ischemic SAEs. The lack of differ-
ence in plasma vasopressin levels between those who did or
did not have an SAE could dissociate the biological plausibility
between SAE and use of vasopressor.
CONCLUSIONS
About 10% of patients with septic shock on vasopressin or
NE infusion had SAEs. The development of SAEs identied
patients with greatly increased (about double) mortality rates
compared with patients who did not have SAEs. In patients
on vasopressin infusion, vasopressin levels were not different
between patients with or without SAEs suggesting that simple
pharmacokinetics and higher vasopressin levels did not cause
SAEs. Patients who had the AA genotype of rs28418396 SNP
had signicantly higher rates of SAEs in both discovery and
validation cohorts; the mechanism of this association requires
further investigation.
REFERENCES
1. Dellinger RP, Levy MM, Rhodes A, et al; Surviving Sepsis Campaign
Guidelines Committee including the Pediatric Subgroup: Surviving
Sepsis Campaign: International guidelines for management of
severe sepsis and septic shock, 2012. Intensive Care Med 2013;
39:165228
2. Obritsch MD, Bestul DJ, Jung R, et al: The role of vasopressin in vaso-
dilatory septic shock. Pharmacotherapy 2004; 24:10501063
3. Obritsch MD, Jung R, Fish DN, et al: Effects of continuous vasopres-
sin infusion in patients with septic shock. Ann Pharmacother 2004;
38:11171122
4. Dnser MW, Mayr AJ, Tr A, et al: Ischemic skin lesions as a compli-
cation of continuous vasopressin infusion in catecholamine-resistant
vasodilatory shock: Incidence and risk factors. Crit Care Med 2003;
31:13941398
5. Russell JA, Walley KR, Singer J, et al; VASST Investigators:
Vasopressin versus norepinephrine infusion in patients with septic
shock. N Engl J Med 2008; 358:877887
6. Hall LG, Oyen LJ, Taner CB, et al: Fixed-dose vasopressin compared
with titrated dopamine and norepinephrine as initial vasopressor ther-
apy for septic shock. Pharmacotherapy 2004; 24:10021012
7. Ferguson-Myrthil N: Vasopressor use in adult patients. Cardiol Rev
2012; 20:153158
8. Annane D, Vignon P, Renault A, et al; CATS Study Group:
Norepinephrine plus dobutamine versus epinephrine alone for
management of septic shock: A randomised trial. Lancet 2007;
370:676684
9. Torgersen C, Dnser MW, Wenzel V, et al: Comparing two differ-
ent arginine vasopressin doses in advanced vasodilatory shock: A
randomized, controlled, open-label trial. Intensive Care Med 2010;
36:5765
10. van Haren FM, Rozendaal FW, van der Hoeven JG: The effect of vaso-
pressin on gastric perfusion in catecholamine-dependent patients in
septic shock. Chest 2003; 124:22562260
11. Holmes CL, Walley KR, Chittock DR, et al: The effects of vasopressin
on hemodynamics and renal function in severe septic shock: A case
series. Intensive Care Med 2001; 27:14161421
12. Phillips KA, Veenstra DL, Oren E, et al: Potential role of pharmacoge-
nomics in reducing adverse drug reactions: A systematic review.
JAMA 2001; 286:22702279
13. Russell JA: Bench-to-bedside review: Vasopressin in the manage-
ment of septic shock. Crit Care 2011; 15:226
14. Nakada TA, Russell JA, Wellman H, et al: Leucyl/cystinyl aminopepti-
dase gene variants in septic shock. Chest 2011; 139:10421049
15. Nakada TA, Russell JA, Boyd JH, et al: Beta2-Adrenergic receptor
gene polymorphism is associated with mortality in septic shock. Am J
Respir Crit Care Med 2010; 181:143149
16. Wu CK, Tsai CT, Chang YC, et al: Genetic polymorphisms of
the angiotensin II type 1 receptor gene and diastolic heart failure.
J Hypertens 2009; 27:502507
17. Nakada TA, Russell JA, Boyd JH, et al: Association of angiotensin II
type 1 receptor-associated protein gene polymorphism with increased
mortality in septic shock. Crit Care Med 2011; 39:16411648
18. Bone RC, Balk RA, Cerra FB, et al: Defnitions for sepsis and organ
failure and guidelines for the use of innovative therapies in sepsis.
Chest 1992; 101:16441655
19. Sutherland AM, Walley KR, Manocha S, et al: The association of inter-
leukin 6 haplotype clades with mortality in critically ill adults. Arch
Intern Med 2005; 165:7582
20. Sibbald WJ, Vincent JL: Round table conference on clinical trials for
the treatment of sepsis. Intensive Care Med 1995; 21:184189
21. Groesdonk HV, Klingele M, Schlempp S, et al: Risk factors for nonoc-
clusive mesenteric ischemia after elective cardiac surgery. J Thorac
Cardiovasc Surg 2013; 145:16031610
22. Hernando F, Schoots O, Lolait SJ, et al: Immunohistochemical local-
ization of the vasopressin V1b receptor in the rat brain and pituitary
gland: Anatomical support for its involvement in the central effects of
vasopressin. Endocrinology 2001; 142:16591668
23. Lolait SJ, Stewart LQ, Jessop DS, et al: The hypothalamic-pituitary-adrenal
axis response to stress in mice lacking functional vasopressin V1b
receptors. Endocrinology 2007; 148:849856
24. Sharshar T, Blanchard A, Paillard M, et al: Circulating vasopressin
levels in septic shock. Crit Care Med 2003; 31:17521758
25. Xu S, Guo S, Jiang X, et al: The role of norepinephrine and nitric oxide
in activities of rat arginine vasopressin neurons in response to immune
challenge. Neurosci Lett 2005; 383:231235
26. Delmas A, Leone M, Rousseau S, et al: Clinical review: Vasopressin
and terlipressin in septic shock patients. Crit Care 2005; 9:212222
27. Leng G, Brown CH, Russell JA: Physiological pathways regulating
the activity of magnocellular neurosecretory cells. Prog Neurobiol
1999; 57:625655