ORIGINAL ARTICLE

Vasopressin and nitroglycerin decrease portal and hepatic

venous pressure and hepato-splanchnic blood flow

n1
E. Wise , K. Svennerholm1, L. S. Bown1, E. Houltz1, M. Rizell2, S. Lundin1 and S.-E. Ricksten1
1
Department of Anaesthesiology and Intensive Care Medicine, Sahlgrenska Academy, University of Gothenburg, Sahlgrenska University Hospital,
Gothenburg, Sweden
2
Department of Transplantation and Liver Surgery, Sahlgrenska Academy, University of Gothenburg, Sahlgrenska University Hospital, Gothenburg,
Sweden

Correspondence Background: Various methods are used to reduce venous blood

E. Wis

en, Department of Anaesthesiology and
Intensive Care Medicine, Sahlgrenska
University Hospital, Bl
a Str
aket 5, SE-413
45 Go€teborg, Sweden.
E-mail: ellinor.wisen@vgregion.se
Conflict of interest
The authors have no conflict of interest.
Registration: Clinical Trials, https://www.
clinicaltrials.gov, NTC02993640.
Submitted 5 February 2018; accepted 23
February 2018; submission 18 December
2017.
Citation
Wise
n E, Svennerholm K, Bown LS, Houltz E,
Rizell M, Lundin S, Ricksten S-E. Vasopressin
and nitroglycerin decrease portal and hepatic
venous pressure and hepato-splanchnic blood
flow. Acta Anaesthesiologica Scandinavica
2018
doi: 10.1111/aas.13117
pressure in the hepato-splanchnic circulation, and hence minimise
blood loss during liver surgery. Previous studies show that com-
bination of vasopressin and nitroglycerin reduces portal pressure
and flow in patients with portal hypertension, and in this study
we investigated this combination in patients with normal portal
pressure.
Method: In all, 13 patients were studied. Measurements were
made twice to confirm baseline (C1 and BL), during vasopressin
infusion 4.8 U/h (V), and during vasopressin infusion combined
with nitroglycerin infusion (V + N). Portal venous pressure
(PVP), hepatic venous pressure (HVP), central haemodynamics
and arterial and venous blood gases were obtained at each mea-
suring point, and portal (splanchnic) and hepato-splanchnic blood
flow changes were calculated.
Results: Vasopressin alone did not affect PVP, whereas HVP
increased slightly. In combination with nitroglycerin, PVP
decreased from 10.1
1.6 to 8.9
1.3 mmHg (P < 0.0001), and
HVP decreased from 7.9
1.9 to 6.2
1.3 mmHg (P = 0.001).
Vasopressin reduced portal blood flow by 47
19% and hepatic
venous flow by 11
18%, respectively. Addition of nitroglycerin
further reduced portal- and hepatic flow by 55
13% and
30
13%, respectively. Vasopressin alone had minor effects on
central haemodynamics, whereas addition of nitroglycerin reduced
cardiac index (3.2
0.7 to 2.7
0.5; P < 0.0001). The arterial-
portal vein lactate gradient was unaffected.
Conclusion: The combination of vasopressin and nitroglycerin
decreases portal pressure and hepato-splanchnic blood flow, and
could be a potential treatment to reduce bleeding in liver resec-
tion surgery.

Editorial comment
This article is accompanied by an editoral.

Blood loss and blood transfusions are suggested Blood loss has been correlated with pressure
to be independent predictors of postoperative within the inferior vena cava6 and central venous
morbidity and mortality after liver surgery.1–5 pressure (CVP) has been used as a surrogate for
Acta Anaesthesiologica Scandinavica (2018)
ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 1

An international journal of anaesthesiology, intensive

care, pain, and critical emergency medicine

ET AL.
E. WISEN
pressures within the liver venous vascular bed.7,8
Various methods are used to reduce blood loss,
such as reducing CVP (low CVP technique) by
fluid restriction, administration of diuretics,
vasodilators and epidural analgesia and surgi-
cally by inflow and outflow occlusion.8–11
To reduce the venous pressures of the hepatic
vascular bed, such as hepatic- (HVP) and portal
venous pressures (PVP), nitroglycerin infusion is
often used during the resection period. Our group
has previously shown that nitroglycerin reduces
PVP and HVP in liver resection surgery,10 and
nitroglycerin has been shown to decrease PVP in
cirrhotic patients with portal hypertension.12 The
use of nitroglycerin is, however, limited by its
systemic vasodilatory effect and the risk of intra-
operative circulatory instability.10
Another way to decrease blood loss in liver
surgery would be to use vasopressin infusion,
which has been proven to reduce PVP in
patients with portal hypertension.13 Also, we
have recently shown that vasopressin in low-to-
moderate doses decreased hepato-splanchnic
and portal blood flow in patients with normal
portal pressure by 14% and 37%, respectively,
but not PVP and HVP.14 Furthermore, vaso-
pressin induced a blood volume centralisation
caused by a constriction of intra-abdominal
capacitance vessels, which could explain why
vasopressin did not affect PVP or HVP despite
the redistribution of blood volume.14
The primary aim of the present study was to
evaluate whether combined treatment with
vasopressin and nitroglycerin could decrease
portal (splanchnic) and hepato-splanchnic blood
flow, and reduce PVP and HVP, in patients
without known portal hypertension, undergoing
liver resection surgery. Secondary aims were to
assess effects on systemic haemodynamics and
lactate production in the splanchnic and hepato-
splanchnic circulation.
Patients and methods
The study was approved by the Regional Ethical
Review Board in Gothenburg (Dnr 879-13, 4
March 2013, amendment T819-16, 21 September
2016). The trial was registered at ClinicalTrials
(https://www.clinicaltrials.gov, NTC02993640).
The inclusion criteria were as follows: (1) elec-
tive open liver resection, (2) American Society
2 ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
of Anaesthesiology (ASA) class I–II, (3) not
more than two antihypertensive medications, (4)
a body mass index < 30 kg/m2 and (5) no
expected anatomical difficulties for catheterisa-
tion, as assessed by the surgeon. After oral and
written consent, data were obtained from 13
patients during the period of September 2016 to
March 2017. Patient characteristics are presented
in Table 1.
In all, 24 patients consented to participate. Of
these, 11 were excluded due to difficulties
regarding the intraoperative measurement proce-
dure observed by the surgeon (n = 8), such as
tumour location close to portal or hepatic veins,
adhesions in the abdomen, difficulties with
catheter placement, over all lengthy procedure,
etc. One patient was discovered to suffer from
recent heart failure (n = 1), and others were
excluded due to logistic reasons (n = 2) such as
a tight operation schedule or incoming emer-
gency surgery. Data were therefore obtained
from 13 patients.
Anaesthetic technique
An epidural catheter was inserted preopera-
tively but not activated until after the experi-
mental procedure. Anaesthesia was induced
with an intravenous bolus of fentanyl 2 lg/kg
and propofol 2–3 mg/kg. Muscle relaxation was
achieved with rocuronium 0.6 mg/kg. Anaesthe-
sia was maintained with sevoflurane in oxygen/
air mixture adjusted to 0.8 MAC and mechanical
ventilation was adjusted to maintain normocap-
nia. Bolus doses of fentanyl and rocuronium
were repeated if needed. After tracheal intuba-
tion, a central venous catheter (Arrow 4-lumen,
7 Fr, 20 cm) was placed in the right internal
jugular vein, and radial and femoral arterial
catheters were inserted. Mean arterial pressure
(MAP, from the femoral arterial line) and CVP
were measured continuously. Cardiac output
was measured using thermodilution (PiCCO;
PULSION Medical System, Feldkirchen, Ger-
many). Patients received infusion of Ringer’s
acetate 1.5 ml/kg/h, as well as bolus doses of
normal saline for PiCCO calibration. Nora-
drenaline was used to maintain MAP between
65 and 70 mmHg during the surgical procedure,
with the infusion rate kept constant throughout
the experimental procedure.
Acta Anaesthesiologica Scandinavica (2018)
 VASOACTIVE DRUGS AND LIVER CIRCULATION

Table 1 Patient demographics.

Patient Age Sex Diagnosis ASA-class Co-morbidity Cardiovascular drugs

1 55 M Cholangiocarcinoma II HT Calcium channel blocker

2 67 M Colon cancer/liver metastases II –
3 43 M Klatskin tumour I –
4 46 M Rectal cancer/liver metastases I –
5 65 M Colon cancer/liver metastases II HT ARB
6 64 M NET/liver metastases II
7 65 M Prostate cancer/liver metastases II ARB, diuretics
8 62 F Colon cancer/liver metastases II –
9 64 M Colon cancer/liver metastases II DM-type I, HT ACE-inhibitor, beta-blocker
10 63 F Cholangial dysplasia I –
11 66 M Rectal cancer/liver metastases I –
12 60 M Rectal cancer/liver metastases II MI, PCI in 2000 Beta-blocker
13 36 M Adenoma I –

M, Male; F, Female; NET, Neuroendocrine tumour; HT, Hypertension; DM, Diabetes mellitus; MI, Myocardial infarction; PCI, Percutaneous
coronary intervention; ARB, Angiotensin II receptor blocker.

measurements were made (measuring point V + N).

Insertion of hepatic and portal venous
catheters
Two catheters (ARROW 16 Ga; Int., Inc., Reading,
PA, USA) were inserted surgically in the portal
and hepatic vein. The tip of the catheter measuring
HVP was located in the hepatic vein outflow
region, 2–3 cm from the inferior caval vein. The tip
of the catheter measuring PVP was placed directly
in the portal vein. The pressure transducers used
for PVP and HVP, as well as MAP and CVP
(CODAN pvb Critical Care GmbH, Forstinning,
Germany), were zeroed and positioned at the right
atrial level. The abdomen was closed temporarily.
Experimental procedure
The experimental procedure and measurements
were performed after dissection of the liver area, but
prior to the liver resection. After a steady-state period
of 10 min, two control measurements were per-
formed (C1 and baseline; BL). Infusion of vaso-
pressin (Argipressin 0.04 U/ml) was then started at
an initial dose of 9.6 U/h for 5 min, followed by
4.8 U/h for 15 min after which measurements were
made (measuring point V). The higher dose was
given to compensate for catheter lumen and ascertain
a rapid achievement of adequate plasma concentra-
tion of vasopressin. Thereafter, an infusion of nitro-
glycerin (Glycerylnitrat 1 mg/ml) was added and
titrated to reach a MAP of 60 mmHg (0.15–2.4 lg/
kg/min). After 5 min of steady state, the last sets of
Acta Anaesthesiologica Scandinavica (2018)
ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
The experimental procedure is illustrated in Fig. 1.
Measurements
At each measuring point, cardiac index (CI),
stroke volume index (SVI), intrathoracic blood
volume index (ITBVI) and systemic vascular
resistance index (SVRI) were obtained from
transpulmonary thermodilution measurements in
duplicate and calculations using the PiCCO soft-
ware (PULSION Medical Systems, Feldkirchen,
Germany), and indexed to body surface area.
Blood samples were obtained from the arterial,
portal, hepatic and central venous catheters at each
measuring point for analysis of blood oxygen
saturation and lactate levels. Two independent
investigators performed the offline analysis of
hepatic pressure recordings, and measurements
were double checked by a third investigator not
involved in performing the original recordings.
Data collection
Patient data were collected using a Philips monitor-
ing system (IntelliVue MX800, Eindhoven, the
Netherlands). PVP and HVP were sampled from the
corresponding pressure catheters by an A/D
converter (BIOPAC MP150, Goleta, CA, USA) at a
sampling rate of 20 Hz and transferred to a software
program (AcqKnowledge; Biopac) for offline
analysis.
3

ET AL.
E. WISEN

Fig. 1. Schematic representation of the experimental procedure. C1,
Control 1; BL, Baseline; V, vasopressin 4.8 U/h; V + N, vasopressin
4.8 U/h + nitroglycerin to a target MAP of 60 mmHg. Vasopressin was
started at 9.6 U/h and then reduced after 5 min.
Statistical analysis
Values are presented as mean
standard devia-
tion (SD). The web-based statistical tool www.
quantitativeskills.com was used for power analy-
sis. To detect a 25% (2–3 mmHg) reduction in
portal pressure, at a SD of 2 mmHg10 with a
power of 0.8 and a two-sided P value of <0.05, a
sample size of nine patients was required. To
detect a 20% relative reduction of portal flow,
with a SD of 15%,14 with a power of 0.8 and a
two-sided P value of <0.05 a sample size of 10
patients was required. To account for missing
data, we aimed to include 13 patients. A two-
way analysis of variance (ANOVA) for repeated
measures was performed. If a significant ANOVA
was present, the analysis was followed by paired
t-tests between BL and V and between BL and
V + N. P values <0.05 were considered signifi-
cant. All analyses were performed in the Statisti-
cal Package for the Social Sciences (SPSS version
23.0, IBM, New York, USA).

Results
Calculation of changes in portal and hepato- Due to technical reasons, HVP measurements
splanchnic venous flow from patient number 2 and CVP measurements
Changes (D) in portal venous flow (Qpv) from patient number 7 were not obtained. Blood
(splanchnic blood flow) and hepatic (Qhv) gas analysis was on two occasions complicated
venous flow (hepato-splanchnic blood flow) rel- by machine failure, and measurement of oxygen
ative to baseline during vasopressin infusion saturation (patients 6 and 13) and hepatic
(V) were calculated using the following equa- venous lactate (patient 6) could not be per-
tion, derived from Fick’s principle:15 formed.
There was no difference between HVP, PVP, sys-
temic haemodynamic parameters, oxygen satura-
DQpvð%Þ
tion values or lactate levels between the two
ðpreVÞ ðpreVÞ
QpvðpostVÞ SaO2  SpvO2 control measurement points (C1 and BL), which
¼ ðpreVÞ
¼ ðpostVÞ ðpostVÞ
Qpv SaO2  SpvO2 confirms a steady state at baseline (BL). The dose of
nitroglycerin required to achieve MAP 60 mmHg
 100; at measuring point V + N was very individual, and
and for hepato-splanchnic flow : DQhvð%Þ ranged between 0.1 and 2.4 lg/kg/min.
ðpreVÞ ðpreVÞ
QhvðpostVÞ SaO2  ShvO2
¼ ¼  100
QhvðpreVÞ ðpostVÞ
SaO2  ShvO2
ðpostVÞ Systemic haemodynamics (Table 2)
Vasopressin increased MAP, SVRI, SVI and
CVP and decreased HR, whereas CI remained
where SaO2 is arterial oxygen saturation, SpvO2 unchanged. As nitroglycerin was added (dosed
is portal venous oxygen and ShvO2 is hepatic individually to achieve a target MAP of
venous oxygen saturation. The same equa- 60 mmHg), SVRI and HR returned to BL values,
tion was used for calculation of flow changes whereas CI, SVI and CVP decreased compared
comparing (V + N) to baseline. to BL. There was a trend towards increased

Acta Anaesthesiologica Scandinavica (2018)

4 ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
 VASOACTIVE DRUGS AND LIVER CIRCULATION

Table 2 Effects of vasopressin and nitroglycerin on systemic haemodynamics.

BL V V+N ANOVA P value

CI (l/min/m2) 3.24
0.67 3.24
0.72 2.70
0.48*** <0.001

MAP (mmHg) 70
7.7 77
6.3** 60
3.5** <0.001
SVRI (dynes 9 s 9 cm5/m2) 1619
378 1801
378* 1653
276 <0.01
ITBVI (ml/m2) 863
307 947
472 884
383 0.12
SVI (ml/m2) 37.2
4.9 38.9
6.4* 30.7
4.4*** <0.001
HR (beats/min) 86
13 83
13** 88
13 0.001
CVP (mmHg) 6.72
1.3 7.53
1.49*** 5.68
1.11** <0.001

BL, baseline; V, vasopressin 4.8 U/h; V + N, vasopressin 48 U/h + nitroglycerin to MAP 60 mmHg; CI, Cardiac Index; MAP, Mean Arterial
Pressure; SVRI, Systemic Vascular Resistance Index; ITBVI, Intrathoracic Blood Volume Index; SVI, Stroke Volume Index; HR, Heart Rate; CVP,
Central Venous Pressure; Values are presented as means
SD; Asterisks indicate significant difference vs. BL. *P < 0.05; **P < 0.01;
***P < 0.001.

ITBVI with vasopressin, which then returned to
baseline after addition of nitroglycerin. The
haemodynamic changes indicate a decreased car-
diac preload with addition of nitroglycerin.
Portal and hepatic pressures (Table 3, Fig. 2)
During vasopressin infusion, 4.8 U/h, PVP
remained unchanged but HVP increased slightly
compared to BL (7.1
1.5 to 7.9
1.9 mmHg
P = 0.001). With nitroglycerin added to on-
going vasopressin, there was a significant reduc-
tion of PVP (10.1
1.6 to 8.9
1.3 mmHg) and
HVP (7.9
1.9 to 6.2
1.3 mmHg), which both
decreased below BL.
Portal and hepatic venous flow (Table 3,
Fig. 3)
Vasopressin infusion decreased portal flow by
47% and hepatic flow by 11% compared to BL.
When nitroglycerin was added, the portal flow
was reduced further, by 55%, and the hepatic
flow by 30%, compared to BL.
Lactate (Table 4)
Lactate levels did not differ between measuring
points in arterial, central venous or portal
venous blood. Hepatic venous lactate increased
at point V + N compared to BL (1.60–
1.82 mmol/l) and the arterio-hepatic lactate gra-
dient decreased (0.32–0.16 mmol/l). The arterio-
portal lactate gradient was not affected.
Discussion
In this study involving patients with normal
portal venous pressure undergoing liver resec-
tion surgery, we showed that the combined
infusion of a low-to-moderate dose of vaso-
pressin and nitroglycerin reduced portal and

Table 3 Venous pressures, oxygen saturation and calculated relative change of portal and hepatic flow.

BL V V+N ANOVA P value

PVP (mmHg) 10.35
1.50 10.05
1.59 8.88
1.25*** <0.001

HVP (mmHg) 7.12
1.40 7.94
1.89** 6.23
1.34*** <0.001
Arterial oxygen saturation (%) 97.7
0.5 97.6
0.7 96.9
1.2** <0.001
Central venous saturation (%) 80.5
3.3 80.8
9 76.2
4.6** <0.001
Portal venous saturation (%) 88.8
2.9 78.4
9 ** 75.8
7.2*** <0.001
Hepatic venous saturation (%) 75.4
10.8 70.4
16.6* 63.7
16.5** <0.001
D portal flow 1 0.53
0.19*** 0.45
0.13*** <0.001
D hepatic venous flow 1 0.89
0.18 0.7
0.13*** <0.001

BL, baseline; V, vasopressin 4.8 U/h; V + N, vasopressin 4.8 U/h + nitroglycerin; PVP, Portal venous pressure; HVP, Hepatic venous pressure;
Values are presented as means
SD. Asterisks indicate significant difference vs. BL at *P < 0.05;**P < 0.01; ***P < 0.001.

Acta Anaesthesiologica Scandinavica (2018)

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 5

ET AL.
E. WISEN
Fig. 2. Vasopressin 4.8 U/h (V) and vasopressin 4.8 U/
h + nitroglycerin (V + N) decrease portal, hepatic venous and central
venous pressure as indicated by the circle, triangle and square-
marked lines, respectively. Asterisks indicate significant differences vs.
BL **P < 0.01; ***P < 0.001.
hepatic venous pressure, and considerably
reduced both portal (splanchnic) and hepato-
splanchnic blood flow. To our knowledge, the
effects of this drug combination on splanchnic
and hepato-splanchnic blood flow and portal
and hepatic venous pressure have not been
tested previously in this clinical setting.
In the present study, vasopressin induced a
constriction of both systemic resistance vessels
and capacitance vessels, inducing an increase in
MAP, as well as a centralisation of blood vol-
ume, reflected by an increase in CVP and SVI.
These findings confirm the results from our
recent study.14 When nitroglycerin was added to
vasopressin, systemic vascular resistance was
normalised compared to baseline, but cardiac
output decreased below baseline. This effect
was likely caused by systemic venodilatation
and a consequent fall in cardiac preload, stroke
volume and cardiac output.Vasopressin infusion
only did not affect portal venous pressure signif-
icantly in the present study, confirming our pre-
vious findings.14 This could be explained by a
vasopressin-induced vasoconstriction and
decreased compliance of abdominal capacitance
vessels, which will offset the fall in intra-
abdominal blood volume, consequently with no
net effect on portal pressure. The lack of effect
6 ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
Fig. 3. Vasopressin 4.8 U/h (V) and vasopressin 4.8 U/
h + nitroglycerin to a MAP of 60 mmHg (V + N) decrease portal and
hepatic venous blood flow, as indicated by the circle and triangle-
marked lines, respectively. Asterisks indicate significant difference vs.
BL ***P < 0.001.
of vasopressin on portal pressure in this study,
on patients with normal portal pressure, is in
contrast to previous studies on patients with
portal hypertension. Westaby et al. showed that
vasopressin decreased portal venous pressure
from 21 to 14 mmHg in patients with portal
hypertension16 and it has also been shown that
portal venous pressures decrease in patients
with portal hypertension undergoing liver trans-
plantation.13,17 When nitroglycerin was added
to vasopressin infusion, the vasoconstrictory
effect of vasopressin was counteracted, resulting
in a fall in both hepatic and portal venous pres-
sure below baseline, as shown in the present
study.
Vasopressin induced a pronounced intestinal
vasoconstriction as reflected by a marked reduc-
tion in portal venous blood flow (47%) in our
study. However, the fall in hepatic venous flow,
reflecting the global hepato-splanchnic perfu-
sion, was less pronounced (11%). The relatively
greater reduction in portal flow compared to
hepatic venous flow is probably due to the hep-
atic arterial buffer response (HABR), a compen-
satory mechanism mediated by adenosine
release, leading to an increased flow of more
Acta Anaesthesiologica Scandinavica (2018)
 VASOACTIVE DRUGS AND LIVER CIRCULATION

Table 4 Lactate and lactate gradients.

BL V V+N ANOVA P value

Arterial lactate (mmol/l) 1.90
0.99 1.96
1.07 1.98
0.96 0.58
Central venous lactate (mmol/l) 2.00
0.94 2.04
0.99 2.11
0.91 0.21
Portal venous lactate (mmol/l) 1.82
0.91 1.88
1.04 1.96
0.94 0.16
Hepatic venous lactate(mmol/l) 1.60
0.98 1.78
1.04* 1.82
1.07* 0.01
Arterio-portal lactate gradient 0.085
0.18 0.083
0.16 0.026
0.24 0.59
Arterio-hepatic lactate gradient 0.32
0.28 0.17
0.31 0.16
0.35* 0.03

BL, baseline; V; vasopressin 4.8 U/h; V + N, vasopressin 4.8 U/h + nitroglycerin; Values are presented as means
SD. Asterisks indicant sig-
nificant difference vs. BL at *P < 0.05.

oxygenated blood in the vascular bed supplied
from the hepatic artery when portal flow
decreases.18–20 When nitroglycerin was added to
vasopressin, portal blood flow and hepato-
splanchnic blood flow were reduced further.
This is explained by the nitroglycerin-induced
fall in MAP, in turn caused by systemic vaso-
and venodilatation, decreasing systemic vascular
resistance, cardiac preload and cardiac output.
Thus, the combined infusion of vasopressin and
nitroglycerin reduced portal blood flow and
hepato-splanchnic blood flow to 55% and 30%
of baseline flow, respectively. In other words,
the combined infusion of vasopressin and nitro-
glycerin decreases both portal venous pressure,
as well as portal inflow of blood to the surgical
field during liver resection surgery.
Different interventions have been suggested to
lower the CVP and reduce bleeding during liver
surgery, aiming to reduce the vascular distend-
ing pressure over the liver parenchyma.7,8 Blood
flow through the liver, and possibly total liver
blood volume is also likely to influence blood
loss during liver surgery. In two animal studies,
vasopressin has been shown to reduce total liver
blood volume.21,22 To surgically control bleed-
ing during liver resection, inflow and outflow
occlusion (e.g. the Hepatic Pedicle Clamping or
Pringle’s Manoeuvre) is used routinely. There
are conflicting results regarding the effect on
outcome and risk of hepatic ischemia using
these methods, as well as risk of increased
haemodynamic instability during clamping.23,24
Our study suggests that a combined infusion of
vasopressin and nitroglycerin could be a phar-
macological alternative to partial occlusion tech-
niques. From a clinical perspective, our method
with a pharmacologically induced reduction of
both pressure and flow is appealing, since it
potentially allows for the HABR mechanism to
better preserve oxygenation of the liver parench-
yma compared to the surgical occlusion tech-
niques. Whether this pharmacological
alternative may reduce bleeding during liver
resection surgery needs to be evaluated in a
prospective randomised trial.
In the present study, the fall in portal venous
and hepato-splanchnic blood flow with the
combined vasopressin and nitroglycerin infu-
sions was accompanied by a minor increase in
hepatic venous lactate and decrease in the arte-
rio-hepatic lactate gradient. This suggests that
this treatment has the potential to induce
splanchnic ischemia and lactate production, and
could be of clinical relevance if this treatment is
to be used for a longer period. On the other
hand, it has also been shown that serum lactate
may increase markedly after use of surgical
occlusion techniques,25 and that this increase is
directly correlated to the cumulative time of
hepatic hilum clamping.26 However, in our
study, the levels of lactate after drug infusions
are still within the normal range, as opposed to
the elevated levels found in studies of mechani-
cal clamping.25
Another concern is that vasopressin in similar
doses as used in the present study may produce
renal vasoconstriction and impair renal oxygena-
tion.27 On the other hand, the use of Terli-
pressin (an analogue of vasopressin) in
patients undergoing liver transplantation
resulted in a lower incidence of acute kidney
injury.28 Furthermore, the fall in splanchnic
blood flow may also have the potential to
induce pancreatic ischemia.29 Thus, in future
prospective studies, not only the efficacy but

Acta Anaesthesiologica Scandinavica (2018)

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 7

ET AL.
E. WISEN
also the safety of combined treatment with vaso-
pressin and nitroglycerin must be addressed.
In the present study, we performed measure-
ments intraoperatively, prior to the liver resection
period, which limited the time available for the
experimental procedure. One could argue
whether a steady state was reached after 15 min
of infusion of each agent. However, plasma half-
life for vasopressin is 1–2 min30 and for nitro-
glycerin 1–4 min, and steady state would be
expected to be achieved approximately after 5–
12 min. Ideally, portal and hepatic venous flow
would have been measured directly and simulta-
neously, but technically this proved to be diffi-
cult. Instead, we chose to calculate the changes in
flow from changes in oxygen content in the portal
vein (draining the splanchnic bed) and in the
hepatic vein, under steady-state conditions,
using Fick’s principle. We assumed that neither
vasopressin nor nitroglycerin themselves affected
splanchnic and hepatic oxygen consumption,
which is a prerequisite for the use of Fick’s prin-
ciple to assess changes in organ flows.
In this study, we did not include a time-con-
trol group. One could therefore argue that the
changes in portal and hepatic venous pressures
and flows were not caused by the investigated
agents themselves, but instead to some extent
by spontaneous fluctuations or time-dependent
effects on these variables. On the other hand,
there was no difference between HVP, PVP, sys-
temic haemodynamic parameters, oxygen satura-
tion values or lactate levels between the two
control measurement points (C1 and BL), which
confirms a steady state before the pharmacologi-
cal intervention. Another limitation is the small
sample size, which may increase the risk of type
II errors.
Conclusion
In the present study on patients with normal
portal pressure undergoing liver resection sur-
gery, we showed that combined infusion of
vasopressin and nitroglycerin reduced portal
and hepatic venous pressure, together with a
substantial fall in both portal (splanchnic) and
hepato-splanchnic blood flow. Whether or not
combined infusion of vasopressin and nitroglyc-
erin can be used to decrease perioperative blood
loss during liver resection surgery, without
8 ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
adverse renal or intestinal effects, has to be
determined in a future randomised, large sam-
ple study.
Acknowledgements
This study was funded by the Swedish State
Support for Clinical Research (ALFGBG-75130),
V€astra G€otaland, Sweden, and the Gothenburg
Medical Society, Gothenburg, Sweden (Grant
number 691901). The authors have no conflict of
interest.
References
1. Wang CC, Iyer SG, Low JK, Lin CY, Wang SH, Lu
SN, Chen CL. Perioperative factors affecting long-
term outcomes of 473 consecutive patients
undergoing hepatectomy for hepatocellular
carcinoma. Ann Surg Oncol 2009; 16: 1832–42.
2. Capussotti L, Muratore A, Amisano M, Polastri R,
Bouzari H, Massucco P. Liver resection for
hepatocellular carcinoma on cirrhosis: analysis of
mortality, morbidity and survival–a European
single center experience. Eur J Surg Oncol 2005;
31: 986–93.
3. Yang T, Zhang J, Lu JH, Yang GS, Wu MC, Yu WF.
Risk factors influencing postoperative outcomes of
major hepatic resection of hepatocellular carcinoma
for patients with underlying liver diseases. World J
Surg 2011; 35: 2073–82.
4. Kooby DA, Stockman J, Ben-Porat L, Gonen M,
Jarnagin WR, Dematteo RP, Tuorto S, Wuest D,
Blumgart LH, Fong Y. Influence of transfusions on
perioperative and long-term outcome in patients
following hepatic resection for colorectal
metastases. Ann Surg 2003; 237: 860–9. discussion
69-70.
5. de Boer MT, Molenaar IQ, Porte RJ. Impact of
blood loss on outcome after liver resection. Dig
Surg 2007; 24: 259–64.
6. Johnson M, Mannar R, Wu AV. Correlation
between blood loss and inferior vena caval pressure
during liver resection. Br J Surg 1998; 85: 188–90.
7. Jones RM, Moulton CE, Hardy KJ. Central venous
pressure and its effect on blood loss during liver
resection. Br J Surg 1998; 85: 1058–60.
8. Melendez JA, Arslan V, Fischer ME, Wuest D,
Jarnagin WR, Fong Y, Blumgart LH. Perioperative
outcomes of major hepatic resections under low
central venous pressure anesthesia: blood loss,
blood transfusion, and the risk of postoperative
Acta Anaesthesiologica Scandinavica (2018)

renal dysfunction. J Am Coll Surg 1998; 187: 620–
5.
9. Wang WD, Liang LJ, Huang XQ, Yin XY. Low central
venous pressure reduces blood loss in hepatectomy.
World J Gastroenterol 2006; 12: 935–9.
10. Sand L, Lundin S, Rizell M, Wiklund J, Stenqvist
O, Houltz E. Nitroglycerine and patient position
effect on central, hepatic and portal venous
pressures during liver surgery. Acta Anaesthesiol
Scand 2014; 58: 961–7.
11. Smyrniotis VE, Kostopanagiotou GG, Contis JC,
Farantos CI, Voros DC, Kannas DC, Koskinas JS.
Selective hepatic vascular exclusion versus Pringle
maneuver in major liver resections: prospective
study. World J Surg 2003; 27: 765–9.
12. Noguchi H, Toyonaga A, Tanikawa K. Influence of
nitroglycerin on portal pressure and gastric mucosal
hemodynamics in patients with cirrhosis. J
Gastroenterol 1994; 29: 180–8.
13. Wagener G, Gubitosa G, Renz J, Kinkhabwala M,
Brentjens T, Guarrera JV, Emond J, Lee HT, Landry
D. Vasopressin decreases portal vein pressure and
flow in the native liver during liver transplantation.
Liver Transpl 2008; 14: 1664–70.
14. Bown LS, Ricksten SE, Houltz E, Einarsson H,
Sondergaard S, Rizell M, Lundin S. Vasopressin-
induced changes in splanchnic blood flow and
hepatic and portal venous pressures in liver
resection. Acta Anaesthesiol Scand 2016; 60: 607–15.
15. Iribe G, Yamada H, Matsunaga A, Yoshimura N.
Effects of the phosphodiesterase III inhibitors
olprinone, milrinone, and amrinone on
hepatosplanchnic oxygen metabolism. Crit Care
Med 2000; 28: 743–8.
16. Westaby D, Gimson A, Hayes PC, Williams R.
Haemodynamic response to intravenous
vasopressin and nitroglycerin in portal
hypertension. Gut 1988; 29: 372–7.
17. Mukhtar A, Dabbous H. Modulation of splanchnic
circulation: role in perioperative management of
liver transplant patients. World J Gastroenterol
2016; 22: 1582–92.
18. Lautt WW. Regulatory processes interacting to
maintain hepatic blood flow constancy: vascular
compliance, hepatic arterial buffer response,
hepatorenal reflex, liver regeneration, escape from
vasoconstriction. Hepatol Res 2007; 37: 891–903.
19. Takala J. Determinants of splanchnic blood flow.
Br J Anaesth 1996; 77: 50–8.
20. Kerr JC, Hobson RW 2nd, Seelig RF, Swan KG.
Vasopressin: route of administration and effects on
Acta Anaesthesiologica Scandinavica (2018)
ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
VASOACTIVE DRUGS AND LIVER CIRCULATION
canine hepatic and superior mesenteric arterial
blood flows. Ann Surg 1978; 187: 137–42.
21. Greenway CV, Lautt WW. Effects of infusions of
catecholamines, angiotensin, vasopressin and
histamine on hepatic blood volume in the
anaesthetized cat. Br J Pharmacol 1972; 44: 177–84.
22. Zhang W, Shibamoto T, Kuda Y, Shinomiya S,
Kurata Y. The responses of the hepatic and
splanchnic vascular beds to vasopressin in rats.
Biomed Res 2012; 33: 83–8.
23. Sanjay P, Ong I, Bartlett A, Powell JJ, Wigmore SJ.
Meta-analysis of intermittent Pringle manoeuvre
versus no Pringle manoeuvre in elective liver
surgery. ANZ J Surg 2013; 83: 719–23.
24. Weiss MJ, Ito H, Araujo RL, Zabor EC, Gonen M,
D’Angelica MI, Allen PJ, DeMatteo RP, Fong Y,
Blumgart LH, Jarnagin WR. Hepatic pedicle
clamping during hepatic resection for colorectal
liver metastases: no impact on survival or hepatic
recurrence. Ann Surg Oncol 2013; 20: 285–94.
25. Pietsch UC, Herrmann ML, Uhlmann D, Busch T,
Hokema F, Kaisers UX, Schaffranietz L. Blood
lactate and pyruvate levels in the perioperative
period of liver resection with Pringle maneuver.
Clin Hemorheol Microcirc 2010; 44: 269–81.
26. Giustiniano E, Procopio F, Costa G, Rocchi L,
Ruggieri N, Cantoni S, Zito PC, Gollo Y, Torzilli G,
Raimondi F. Serum lactate in liver resection with
intermittent Pringle maneuver: the “square-root-
shape. J Hepatobiliary Pancreat Sci 2017; 24: 627–
36.
27. Bragadottir G, Redfors B, Nygren A, Sellgren J,
Ricksten SE. Low-dose vasopressin increases
glomerular filtration rate, but impairs renal
oxygenation in post-cardiac surgery patients. Acta
Anaesthesiol Scand 2009; 53: 1052–9.
28. Mukhtar A, Mahmoud I, Obayah G, Hasanin A,
Aboul-Fetouh F, Dabous H, Bahaa M, Abdelaal A,
Fathy M, El Meteini M. Intraoperative terlipressin
therapy reduces the incidence of postoperative
acute kidney injury after living donor liver
transplantation. J Cardiothorac Vasc Anesth 2015;
29: 678–83.
29. Krejci V, Hiltebrand LB, Jakob SM, Takala J,
Sigurdsson GH. Vasopressin in septic shock: effects
on pancreatic, renal, and hepatic blood flow. Crit
Care 2007; 11: R129.
30. Janaky T, Laszlo FA, Sirokman F, Morgat JL.
Biological half-life and organ distribution of [3H]8-
arginine-vasopressin in the rat. J Endocrinol 1982;
93: 295–303.
9