vascular effects of loop diuretics - cardiovascular research

GmdiovascularResearch

Cardiovascular Research 32 (1996) 988-997

Review

Vascular effects of loop diuretics

Tom P.J. Dormans, Peter Pickkers, Frans G.M. Russel, Paul Smits *Department ofPhurmacologj, University Hospital, PO Box 9101, 6500 HB Nijrnegen,Netherlands

Received 2 April 1996; accepted 21 May 1996

Abstract

Although it is generally believed that the beneficial effect of loop diuretics is the result of a rapid increase in diuresis, substantialevidence, from a large number of in vitro and in vivo experiments, has accumulated showing that administration of furosemide causesdirect vascular effects, which probably contribute to its acute clinical effects. Several mechanisms are involvedin the vascular response toloop diuretics. The role of the renin–angiotensin–aldosterone axis, prostaglandins and the direct vascular effects of loop diuretics on boththe arterial and venous parts of the vasctdature are discussed.

Keword.$: Loop diuretic; Fumsemide; Venodilation; Prostaglandin; Angiotensin II; Salt depletion

1. Introduction

Diuretic therapy has proved to be effective in thetreatment of acute and chronic heart failure. The potentloop diuretics, furosemide and bumetanide, are frequentlyused in the treatment of disease states characterized byfluid and sodium retention. After intravenous administra-tion of furosemide, clinical relief of symptoms often pre-cedes the increase in diuresis in patients with acute heartfailure, suggesting the presence of an extrarenal effect.Although it is generally believed that the beneficial effectof loop diuretics is the result of a rapid increase in diuresis,substantial evidence, from a large number of in vivo and invitro experiments, has accumulated showing that adminis-tration of furosemide causes vascular effects, which proba-bly contribute to its acute clinical effects.

At first sight the reports on the vascular non-diureticeffects of furosemide seem conflicting. However, a greatdeal of the disparity in the results seems to be due todifferences in the vascular bed studied (arterial or venous,renal or pulmonary, etc.), the species studied, the timing(acute vs. chronic effects), systemic vs. local effects, directvs. indirect effects and differences in disease states. In thispaper the literature on vascular effects of loop diuretics

(with emphasis on furosemide) is reviewed with referenceto the differences in experimental protocols. Finally, somegeneral conclusions are drawn, and suggestions for futureinvestigations are given.

2. In vitro studies

The direct vascular effects of furosemide are difficult tostudy in vivo because of interfering counteracting mecha-nisms which may even completely mask direct effects. In anumber of in vitro studies the presence of Na+K+Cl -co-transport activity has been demonstrated in endothelialas well as vascular smooth muscle cells [1–3], and thisobservation represents a primary focus of interest withregard to the vascular effects of furosemide. However,inhibition of Na+K+Cl - co-transport activity occurs onlyat high furosemide concentrations. These concentrationsare reached in the renal tubule, but not in the cardio-vascular system [4]. It should be emphasized that in all invitro studies much higher concentrations were needed toinduce vascular responses than in the human in vivosituation. An additional difference causing much higherconcentrations of free furosemide is the absence of proteinbinding in the media used.

“ Corresponding author. Tel.: (+31-24) 361391; fax: (+31-24)3614214. Timefor primary review 21 days.

0008-6363/96/$ 15.00 Copyright 0 1996 Elsevier Science B.V. All rights reserved.PI/ S0008-6363(96)00 134-4

Dow

nloaded from https://academ

ic.oup.com/cardiovascres/article/32/6/988/319823 by guest on 26 N

ovember 2021

.5 .*

Tab

le1

5

Invi

tro

expe

rim

ents

onth

eva

scul

aref

fect

sof

furo

sem

ide

1’V

asct

rlat

ure

Spe

cies

Dru

gco

ncen

trat

ion

Mai

nef

fect

sR

efer

ence

~

Art

eria

lm

Aor

taR

abbi

t33

0J.

Lg/

rnl

nH

yper

pola

risa

tion

and

rela

xati

onof

smcd

lrm

uscl

ece

lls

[6]

7M

esR

at1–

81p,

g/m

lD

ose-

depe

rrde

ntin

hibi

tion

ofre

spon

seto

nore

pine

pbri

neU

’]p

Err

rand

rena

lR

abbi

t2o

p,g/

rnl

Irrh

ibit

iono

fre

spon

seto

nore

pine

pbri

rre,

atte

nuat

edby

albu

min

b]g

Ear

Rab

bit

20–3

30pg

\ml

Dir

ecte

ndot

heli

um-i

ndep

ende

ntre

laxi

ngef

fect

p]g

pu~m

es,

tib

Dog

32–9

60~g

/ml

No

rela

xati

onof

prec

orrt

ract

edar

teri

esQ

[10]

zT

ail

Rat

5m

g/kg

*E

ndot

heli

um-d

epen

dent

redu

ctio

nof

elec

tric

ally

stim

ulat

edco

ntra

ctio

n=

[11]

:

Ven

ous

&

Port

alz

Rat

1-10

0pg

/rnl

Red

ucti

onof

resp

onse

tono

repi

neph

rirr

eand

angi

oten

sin

11[5

]n

Port

alR

abbi

t20

*g/m

lg

Inhi

biti

onof

resp

onse

tono

repi

neph

rine

,att

enua

ted

byal

bum

in[8

]a-

Pul

mon

ary

Dog

32-9

60pg

/rnl

End

othe

lium

-ind

eper

rden

trela

xati

onof

nore

pine

phri

n-in

duce

dco

ntra

ctio

nb

[10]

N 2 ,..

Ear

=ce

ntrr

dea

rar

tery

;Mes

=m

esen

teri

c;Pu

l=pu

lmon

ary;

Tib

=an

teri

ortib

ial

arte

ry.

&

*A

dmin

iste

red

dose

inst

ead

offu

rose

mid

eco

ncen

trat

ion.

w 03 co w m a

Dow



ovember 2021

990 T.P.J. Dormans et al. / Cardiovascular Research 32 (1!?96)988–997

The in vitro studies focussing on the vascular effects offurosemide are summarized in Table 1. In the early 1970’san inhibitory effect of furosemide on the vasoconstrictorresponse to norepinephrine and angiotensin H was ob-served in the rat portal vein [5]. It was demonstrated thatincubation with furosemide causes a membrane hyperpo-larisation of 5.5 mV in the relaxed rabbit pulmonary artery[6]. Many vasodilatory agents act by hyperpolarisation ofthe plasma membrane and subsequent closure of voltage-dependent calcium channels, so this observation is consis-tent with, and possibly explains, the direct vasodilatoryaction of furosemide.

Furosemide appeared to have a direct vascular effect inthe perfused mesenteric vascular bed of the rat [7]. In an invitro study with arterial vascular smooth muscle in seg-ments of rabbit blood vessels, furosemide (20 pg/ml)induced a small decrease in resting tension [8].

In the isolated rabbit central ear artery a direct relaxingeffect of furosemide on isolated vessel segments wasconcentration-dependent (O.1–1.0 mM furosemide) [9]. Itwas demonstrated that inhibition of Na+ K+C1 co-trans-port activity or hyperpolarization of the membrane wasunlikely to be the sole mechanism responsible for thevasorelaxant effect of furosemide.

In an in vitro study using dogs it was demonstrated thatfurosemide did not have a direct effect on arterial smoothmuscle, but exhibits selective venorelaxant activity [10].The magnitude of this effect was most pronounced in thepulmonary vascular bed. Moreover, the vasorelaxant activ-ity of furosemide was independent of endothelium, nitricoxide, cyclic GMP and prostanoids.

The role of the endothelium in the direct vasculareffects of furosemide is still unclear. Whereas one reporton an ex vivo experiment showed that the effect offurosemide on the response to sympathetic stimulation wasendothelium-dependent [11], others did not find an impor-tant role for the endothelium in mediating the relaxationcaused by furosemide in vitro [9]. The discrepancy be-tween these results with respect to the endothelium-de-pendency may be caused by the different concentrations offurosemide studied and by the use of albumin-containingsolutions [8].

3. In vivo studies after systemic administration

During the 1970’s interest increased in the vasculareffects of diuretics. With the development of tools tomonitor changes in haemodynamic parameters, these ef-fects could be described more appropriately. In most ofthese studies, as discussed in the next paragraph andsummarized in Table 2, loop diuretics were administeredsystemically. However, it should be noted that the changesin haemodynamic parameters observed directly after ad-ministration of the loop diuretic do not necessarily implydirect vasoactivity of the loop diuretic.

A study by Dikshit et al. is one of the first reports thatfocussed on the vascular effects of loop diuretics [12]. In20 patients with left ventricular failure, intravenous admin-istration of furosemide caused a prompt fall in left ventric-ular filling pressure, which was accompanied by an in-crease in venous compliance, the latter being a marker forvenodilatation. These phenomena preceded an increase inurine and electrolyte output. In dogs, furosemide produceda rapid reduction in pulmonary wedge pressure and anincrease in venous compliance even though the ureterswere ligated [13]. These observations indicate that thisvenous effect may not have been the result of a decrease inplasma volume. The dissociation of diuretic and vasculareffects was confirmed in a study with hypertensive pa-tients: despite a fall in blood pressure, plasma volume didnot change after administration of furosemide in combina-tion with a high salt intake [14]. In patients with peripheraledema and mild hypertension the use of furosemide re-sulted in a decrease in mean arterial pressure, cardiacoutput and total peripheral resistance, whereas the venouscapacitance increased without change in plasma and bloodvolume [15]. However, the dissociation between venodila-tion and plasma volume is not always obvious. In patientswith mild heart disease or hypertension, 80 mg iv.furosemide caused a decrease in right atrial pressure, pul-monary arterial pressure and pulmonary artery wedge pres-sure (signifying increased venous compliance), togetherwith a decrease in cardiac index within 20 min [16]. In thisstudy, a haemoconcentration was observed as well, sug-gesting that the haemodynamic effects were secondary tointravenous volume reduction through diuresis.

The relationship between haemodynamic and hormonalchanges after furosemide injection and during chronicfurosemide treatment was studied in patients with conges-tive heart failure [17]. Cardiac output decreased signifi-cantly after furosemide injection (1 mg/kg body weight),reached its nadir after 90 min and returned to baselinewithin 4 h. The mean pulmonary arterial pressure de-creased steadily throughout the 4 h observation period.These changes were not accompanied or preceded bychanges in plasma renin activity, angiotensin 11or aldos-terone. In this study patients were on a fixed diet; urinelosses were not replaced isovolumetrically. After continu-ous oral furosemide therapy during 8–10 days reciprocalchanges between haemodynamic and hormone indices wereobserved. As the diuretic response diminished, cardiacoutput and pulmonary arterial pressure declined, whereasthe renin–angiotensin system was activated. This suggeststhat during chronic therapy plasma renin activity andangiotensin II might counteract the vasodilatory effects offurosemide. However, there are some reports that are indisagreement with this hypothesis [18,19]. In patients withsevere congestive heart failure, intravenously administeredfurosemide caused an early fall in stroke volume index anda quick transient increase in the systemic vascular resis-tance, a rise in mean arterial blood pressure (within 20 min

Dow



ovember 2021

Tab

le2

Invi

voex

peri

men

tson

the

acut

eha

emod

vnam

icef

fect

sof

furo

sem

ide

Spe

cies

(dis

ease

stat

e)D

ose

Mai

nef

fect

sR

efer

ence

Hum

an(A

HF

)D

og(l

igat

edur

eter

s)H

uman

(hyp

erte

nsio

n)H

uman

(hyp

erte

nsio

n)H

uman

(hyp

erte

nsio

n)H

uman

(CH

F)

Hum

an(C

HF

)R

at(h

yper

tens

ion)

Hum

an(s

altd

eple

ted)

0.5–

l.0m

g/kg

i.v.

2mg/

kg12

0–40

00m

g/da

yp.o

.12

0–20

0m

g/da

yp.

o.80

mg

iv.

1m

g/kg

1.3

+0.

6m

g\kg

iv.

3m

g/kg

iv.

5–80

mg

iv.

Dec

reas

ePW

P,in

crea

seV

C,M

AP

and

CO

unch

ange

dD

ecre

ase

PWP,

PAP,

incr

ease

VC

,SV

R,M

AP

unch

ange

dD

ecre

ase

MA

P,bl

ood

volu

me

unch

ange

dD

ecre

ase

MA

Pan

dSV

R,i

ncre

ase

VC

,blo

odvo

lum

eun

chan

ged

Dec

reas

ePA

Pan

dPW

Pan

dbl

ood

volu

me,

VC

fore

arm

unch

ange

dD

ecre

ase

PAP

and

CO

Dec

reas

eSV

I,in

crea

seM

AP

and

SVR

Dec

reas

eC

Oan

dSV

I,in

crea

seSV

RIn

crea

seV

C

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

AH

F=

acut

ehe

artf

ailu

re;C

HF

=ch

roni

che

artf

ailu

re;C

O=

card

iac

outp

ut;i

v.=

intr

aven

ousl

y;M

AP=

mea

nar

teri

alpr

essu

re;P

AP=

pulm

onar

yar

teri

alpr

essu

re;p

.o.=

oral

ly;P

WP

=pu

lmon

ary

wed

gepr

essu

re;

SVI

=st

roke

volu

me

inde

x;SV

R=

syst

emic

vasc

ular

resi

stan

ce;

SW=

stro

kevo

lum

ein

dex;

VC

=ve

nous

capa

cita

nce.

w w

Dow



ovember 2021

w Q FJ

Tab

le3

m

Exp

erim

ents

onth

ero

leof

the

kidn

eyin

the

vasc

ular

effe

cts

offu

rose

mid

e

. Q <S

peci

es(d

isea

sest

ate)

Dos

eM

ain

effe

cts

Ref

eren

cep

Hum

an(f

unct

iona

lly

anep

hric

)3m

g/kg

i.v.

Incr

ease

FBF,

unch

ange

dSB

P,V

C,w

eigh

t,he

mat

ocri

t,PR

A[2

1]~

Hum

an(a

neph

ric)

80m

giv

.$

VC

and

BP

unch

ange

d[2

2]~

Rat

(acu

tene

phre

ctom

y)5

mg\

kgiv

.C

ompl

ete

inhi

biti

onof

vaso

cons

tric

torr

espo

nse

toN

Ean

dA

T11

[23]

s

Dog

(lig

atio

nur

eter

s)2

mg/

kgiv

,D

ecre

ase

PAP

and

PWP,

incr

ease

SVR

and

VC

,BP

unch

ange

d[1

3]:

Dog

(acu

tene

phre

ctom

y)2

mg/

kgiv

.N

oha

emod

ynam

icch

ange

s[1

31% 2 2

AT

II=

angi

oten

sin

H;

BP=

bloo

dpr

essu

re;

FB

F=

fore

arm

bloo

dfl

ow;

i.v.=

intr

aven

ousl

y;N

E=

nore

pine

phri

n;PA

P=pu

lmon

ary

arte

rial

pres

sure

;P

RA

=pl

asm

aren

inac

tivi

ty;P

WP=

pulm

onar

y1.-

wed

gepr

essu

re;

SBP

=sy

stol

icbl

ood

pres

sure

;SV

R=

syst

emic

vasc

ular

resi

stan

ce;

VC

=ve

nous

capa

cita

nce.

b u r

Dow



ovember 2021

T.P.J. Dormans et al. /Cardiovascular Research 32 (1996) 988–997 993

after injection), associated with an increase in plasma reninactivity, norepinephrine and plasma arginine-vasopressinlevels [18]. These results were strengthened by the out-come of a study in spontaneously hypertensive rats [19] inwhich furosemide (3 mg/kg) caused an early fall in strokevolume and cardiac index. A decrease in mean arterialblood pressure was observed after a delay of 2 to 4 h,which was sustained for 6 to 8 h after injection. Totalperipheral vascular resistance increased substantially andreturned to baseline range within 24 h. The supposedmechanisms involved in the differences between the acuteand chronic effects include an adaptation of baroreflexactivity, a direct vasodilatory effect of diuretics, a de-creased reactivity of the vascular system to pressor stimuli,a reduction of extracellular body fluid volume, and/or theproduction of endogenous vasodilator substances mediatedby furosemide.

The dose-dependency of the vascular effects offurosemide was characterized in healthy volunteers byusing dosages ranging from 5 to 80 mg [20]. Increases invenous capacitance were observed 5 min after iv. adminis-tration of 5 and 10 mg furosemide. Over the dose range20–80 mg, no significant increases were observed. How-ever, after 10 min venous responses showed significantincreases in venous capacitance, equally for all dosagesused. An oral dosage of 80 mg furosemide produced a risein venous capacitance 15 min after administration and adecrease in forearm bloodflow 15–60 min after administrat-ion. A decrease in calf blood flow was observed within 15min following administration of furosemide, regardless ofsalt balance or use of indomethacin [20]. This latter effectof furosemide was associated with a rise in plasma reninactivity and was not observed in anephric patients [21].

4. The role of the kidneys

In an attempt to elucidate the role of the kidneys in thehaemodynamic effect of furosemide, vascular responseswere studied in functionally anephric hypertensive patients[21]. In contrast to experiments in subjects with normalrenal function, intravenously administered furosemidecaused a significant increase in forearm blood flow of 55%within 15 rein, whereas venous capacitance, weight hemat-ocrit and plasma renin activity were unchanged (see Table3). Possibly, this represents a direct vascular effect offurosemide, which becomes unmasked in the absence ofcounteracting mechanisms, such as the renin–angiotensinsystem. In another study, the effect of intravenously ad-ministered furosemide on venous capacitance and calfblood flow was compared in healthy volunteers andanephric patients [22]. Venous capacitance increased inhealthy volunteers, but not in anephric patients. Moreover,this effect of furosemide required a salt-retaining state andit could be blocked by the use of the cyclo-oxygenaseblocker, indomethacin, suggesting an important role for

renal prostaglandins in the systemic vascular effects offurosemide.

Furosernide (5 mg/kg) attenuated the vasoconstrictorresponses of the mesenteric blood vessels in the rat to bothexogenous angiotensin 11 and norepinephrine [23]. Acutebilateral nephrectomy or treatment with indomethacin (2mg/kg iv.) completely prevented this inhibitory effect. Ina subsequent report the inhibitory effect of furosernide onthe vasoconstrictor response to sympathetic nerve stimula-tion was absent after chemical renal medullectomy [24].The authors explained this effect by postulating that in therenal medulla non-prostanoid vasodilatory lipids are pro-duced which mediate the vasodilatory effect of furosemide[25]. Intrarenal prostaglandins probably are involved in therelease of such a lipid. Although substantial evidence of adirect vascular effect of furosemide is available fromseveral in vitro experiments (see foregoing and Table 1), acoincidence of hormonal changes with the observed vascu-lar effects was not considered.

5. The renin-angiotensin-aldosterone system

The release of renin is controlled by three mechanisms:the intrarenal baroreceptor, the sympathetic nervous sys-tem and the macula densa receptor [26]. Results of somestudies show a participation of prostaglandins in reninrelease [27–29]. It was demonstrated that prostaglandinsmediate renin release in response to intrarenal baroreceptorstimulation [30]. On the other hand, renin release due tosymthathetic nerve stimulation is prostaglandin-independent [31]. Micropuncture experiments in rats indi-cate that renin release resulting from macula densa recep-tor stimulation during sodium deprivation is prosta-glandin-dependent [29], whereas in dogs the macula densamechanism of renin release could be blocked by inhibitionof prostaglandin synthesis [32]. It is known from ex vivoexperiments that furosemide exerts a direct stimulatingeffect on renin secretion [33]. In the isolated perfused ratkidney, furosemide-stimulated renin secretion did not re-quire intact PG12 synthesis [34]. The authors proposed thatincreased prostaglandin production and increase of reninrelease after furosemide administration is not causallyrelated, but may be based on a common response tochanges in sodium balance. In fact, prostaglandin synthesiscould even be a counteracting mechanism participating inthe vasoconstrictor action of angiotensin H [35].

The importance of angiotensin II in the vascular effectsof 5 mg intravenously administered furosemide was stud-ied in healthy volunteers [36]. Captopril 50 mg abolishedthe acute increases in venous capacitance and attenuatedthe increase in forearm vascular resistance. The mecha-nism suggested is that angiotensin 11is formed secondarilyto furosemide-stimulated renin release, and that the de-crease in forearm blood flow is the result of the vasocon-strictive effect of angiotensin II. Angiotensin II receptors

Dow



ovember 2021

994 T.P.J. Dormans et al. / Cardiovascular Research 32 (1996) 988-997

are virtually absent in veins, so the net effect appears to bevenodilation due to the angiotensin-induced release ofvasodilatory prostaglandins from the kidney [36]. Thisview may not be entirely correct, as it has been demon-strated that angiotensin II has a direct venoconstrictiveeffect on the human dorsal hand vein [37].

To determine whether the vascular effects of furosemideare shared by bumetanide, another frequently used loopdiuretic, the vascular and renal effects of equipotentdosages of furosemide and bumetanide were compared inhealthy volunteers with moderate [38] and severe saltdepletion [39]. In the case of moderate salt depletion, bothfurosemide (10 and 100 mg) and bumetanide (250 kg and250 mg) caused an increase in renal blood flow in bothdosages. Changes in peripheral vascular responses did notdiffer from placebo. Both treatments led to an acute in-crease in urinary prostaglandin metabolize excretion (whichmay be a reflection of an increased renal blood flow) andplasma renin activity (the latter not increased bybumetanide 250 Kg). Angiotensin II was increased signifi-cantly 30 min after 100 mg furosemide and 2.5 mgbumetanide. Plasma norepinephrine was not influenced byany of the treatments [38]. In contrast with these observa-tions was the vascular response to furosemide (10 and 20mg) and bumetanide (250 and 500 pg) in marked saltdepletion [39]. Significant reductions in forearm bloodflow were observed after both furosemide dosages, but notafter either of the bumetanide dosages. Both drugs had nosignificant influence on venous capacitance. Furosemideinduced an increase in plasma renin activity, whereasbumetanide did not. The differences between furosemideand bumetanide with regard to acute arterial vasoconstric-tive activity may be attributed to the ability of furosemide(and the disability of bumetanide) to stimulate acute reninrelease from the kidney.

The discrepancy between the results of this study [39]and that by Johnston et al. [38] with respect to vasculareffects may be caused by differences in the degree of saltdepletion. This is emphasized by others [17,40]. There areno in vitro studies that compare the vascular effects offurosemide and bumetanide.

As illustrated in the foregoing paragraphs, the totalbody sodium content is an important factor in the modula-tion of the indirect vascular response to furosemide. Ad-ministration of a loop diuretic to a salt-depleted subjectmay further activate the renin–angiotensin system, causinga more pronounced arterial vasoconstriction.

6. Prostaglandins

In 1975 it was shown in dogs that pretreatment with thecyclo-oxygenase inhibitor indomethacin blocked the in-crease in renal blood flow caused by furosemide [41].Since then several studies have explored the role of prosta-glandins in the natriuretic and vascular responses to

furosemide [30,34,40,42–52]. It is of importance to distin-guish the effects of circulating prostaglandins of renalorigin from prostaglandins produced in the local (ex-trarenal) vascular bed, since the furosemide-induced vascu-lar effects may well be dependent on prostaglandins lo-cally produced in the vessel wall. However, in in-vivoexperiments it is difficult to study these two sources ofprostaglandins separately.

The kidney releases PGIZ, PGEZ, PGFza and throm-boxane Az [42]. PGIZ and PGEZ possess important va-sodilatory properties under conditions of prior vasocon-striction. Prostaglandin-induced vasodilatation plays an im-portant role in the maintenance of glomendar filtration andperfusion by dilatation of the afferent arteriole in a salt-de-pleted state, when the renin–angiotensin system is acti-vated [53].

Furosemide has been shown to increase the urinaryexcretion of prostaglandin [38,50,52]. Whether this iscaused by increased renal blood flow or by increasedproduction of prostaglandins is unclear. On the other hand,reports on the effects of inhibition of prostaglandin synthe-sis on furosemide-induced natriuresis are conflicting, prob-ably due to variations in salt balance during the experi-ments [42].

In healthy volunteers PGIZ induced renin release andfurosemide-induced renin release were associated with re-nal PGIZ formation [54]. In a study performed in nor-motensive volunteers, indomethacin (75 mg) decreasedboth the peak urine flow rate and total sodium excretionwithin 1 h of a 30 mg iv. furosemide dose, while anincrease in renal plasma flow and glomerular filtration rateafter furosemide was inhibited [48]. The increase in urinaryexcretion of PGE2 was abolished by indomethacin. Theurinary excretion of a metabolize of systemic PGIZ wasunaltered after furosemide injection. The authors statedthat the early haemodynamic effects of furosemide dependon an increased synthesis of prostaglandins, particularlyPGEZ and probably also PG12. However, it is questionablewhether the non-renal effects are a result of increasedcirculating prostaglandin levels [24,43]. Arguments thatunderscore these doubts are: prostaglandins are very labile,are rapidly metabolized, and increased plasma levels ofprostaglandins have never been measured after furosemideadministration.

Although the studies mentioned above suggest thatfurosemide induces an increment in renal prostaglandinproduction, they do not clarify whether systemic prosta-glandin synthesis—the local production in the extrarenalvasculature—is increased by furosemide. Mediation of thecardiovascular effects of furosemide by vascular productsof arachidonate metabolism were studied in ex vivo experi-ments using an isolated perfused canine lung lobe [47].Furosemide decreased the mean pulmonary artery pressure.This direct arterial vasodilatory activity of furosemide wassimilar to that of PGIZ and could be inhibited with indo-methacin, suggesting that furosemide induces a local pro-

Dow



ovember 2021

T.P.J. Dorrrrarsset al. /Cardiovascular Research 32 (1996) 988-997

Furosemide

/

Vasodilation?

JResistance

artery

‘r

Vascular Via inhibition ofprostaglandin

synthesis? NaKC1-cotransport?

JJ\

Vasodilation Direct effects

*\

on kidney

Capacitance Renal prostaglandinsynthesis

veinRenal PRA synthesis

f’ Lessor no /

995

Vasoconstriction

\’4n9!~ns7.ngiotensin I

Fig. 1. Diagrammatic scheme illustrating the possible mechanisms of furosemide-induced vasoactivity on artery and vein. The upper part of the figurerepresents direct vascular effects, whereas the lower part represents the vascular effects mediated by hormonal changes.

duction of PG12 in resistance and/or capacitance vessels.Recently, an in vitro study was published showing thatfurosemide in primary cultured bovine aortic endothelialcells stimulated the formation of endothelium-derivedkinin, a potent stimulator of endothelial nitric oxide andPG12 formation [43]. These experiments suggest thathaemodynamic effects of furosemide are mediated byprostaglandins released from the local vasculature.

7. Conclusions

Although in the past 25 years much research has beendone, the exact mechanism by which furosemide inducesits vascular effects remains unclear. In Fig. 1 the mecha-nisms involved in the vascular effects are shown. It seemsclear that both direct and indirect mechanisms play a role.The venous vascular response to furosemide appears to bea direct effect, while the arterial response in vitro onlyoccurs at supratherapeutic concentrations, and probably ismediated and modified by other factors like the degree ofsalt depletion, renin, angiotensin II and prostaglandins invivo. The prostaglandins are either produced by the kid-neys or by the endothelium, whereas the precise role of theendothelium has not yet been completely clarified.

Much attention has been paid to the arterial response,while the effects on the venous component have only beenroughly monitored due to a lack of sensitive techniques tomonitor local venous effects. However, especially in pa-tients suffering from cardiac failure, the venous vasodila-tion might be of importance in the observed acute benefi-cial effects.

There are two methods available to study direct vascu-lar effects in vivo. First, direct effects on resistance arteriesin the human forearm can be studied with the perfusedforearm technique. Using this method, direct vasoconstric-tive or vasodilator effects on resistance arteries in thehuman forearm can be examined by drug administrationinto the brachial artery and venous occlusion plethysmo-graphic recordings [55]. Second, with a linear variabledifferential transducer it is possible to measure directvenous vascular effects on a selected dorsal hand vein[56,57]. With these methods it is possible to examine localvascular effects without provoking systemic counterregula-tory effects. In a quest to explore the genuine directvascular effects of loop diuretics in vivo, these methodsprovide the best options for future studies.

References

[1] O’Donnell ME, Owen NE. Regulation of ion pumps and carriers invascular smooth muscle. Physiol Rev 1994;74:683–721.

[2] O’Donnell ME, Owen NE. Sodium cotransport in vascular smoothmuscle cells. Blood Vessels 1991;28:138–146.

[3] O’Donnell ME. [3H]Bumetanide binding in vascular endothelialcells. Quantitation of Na-K-Cl cotransporters. J Biol Chem1989;264:20326–20330.

[4] Ellory JC, Stewart GW. The human erythrocyte Cl-dependent Na-Kcotransport system as a possible model for studying the action ofloop diuretics. Br J Phnrrnacol 1982;75:183-188.

[5] Blair West JR, McKinley MJ, McKenzie JS. Effect of frusemide onthe reactivity of rat portal vein. J Pharm Pharmacol 1972;24:442–446.

[6] Kreye VA, Bauer PK, Villhauer I. Evidence for furosemide-sensitiveactive cbloride transport in vascular smooth muscle. Eur J Pharma-CO1 1981;73:91–95.

Dow



ovember 2021

996 iCP.J.Dormans et al. /Cardiovascular Re.rearch32 (1996) 988-997

[7] Handa M, Kondo K, Saruta T. Effects of diuretics on the vasocon-strictor responses to norepinephrine and potassium ions in the ratmesenteric artery. Arch Int Pharmacodyn Ther 1983;262:124–131.

[8] Andreasen F, Christensen JH, The effect of furosemide on vascularsmooth muscle is influenced by plasma protein. Pharmacol Toxicol1988;63:324–326.

[9] Tian R, Aalkjaer C, Andreasen F. Mechanisms behind the relaxingeffect of furosemide on the isolated rabbit em artery, PharmacolToxicol 1991;68:406–410,

[IO] Greenberg S, McGowan C, Xie J, Summer WR. Selective pul-monary and venous smooth muscle relaxation by furosemide: acomparison with morphine, J Pharmacol Exp Ther 1994;270:1077–1085,

[11] Gerkens JF. Inhibitory effect of frusemide on sympathetic vasocon-strictor responses: dependence on a renal hormone and the vascularendothelium. Clin Exp Pharmacol Physiol 1987;14:371–377.

[12] Dikshit K, Vyden JK, Forrester JS, Chatterjee K, Prakash R, SwanHJ. Renal and extrarenal hemodynamic effects of’ furosemide incongestive heart failure after acute myocardial infarction. N Engl JMed 1973;288:1087–1090.

[13] Bourland WA, Day DK, Williamson HE. The role of the kidney inthe early nondiuretic action of furosemide to reduce elevated leftatria] pressure in the hypervolemic dog. J Pharmacol Exp Ther1977;202:221–229.

[14] Cantarovich F, Benedetti L, Fernandez JC, et al. High doses offurosemide and sodium in hypertension. A single therapeutic mea-sure in severe cases. Nephron 1974;12:133–139.

[15] Niarchos AP, Magrini F. Hemodynamic effects of diuretics inpatients witb marked peripheral edema and mild hypertension. ClinPharmacol Ther 1982;31:370–376.

[16] Hesse B, Nielsen I, Lund Jacobsen H. The early effects of intra-venous frusemide on central haemodynamics, venous tone andplasma renin activity. Clin Sci Mol Med 1975;49:551-555.

[17] Ikram H, Chan W, Espiner EA, Nicholls MG. Haemodynamic andhormone responses to acute and chronic frusemide therapy in con-gestive heart failure. Clin Sci 1980;59:443-449.

[18] Francis GS, Siegel RM, Goldsmith SR, Olivari MT, Levine TB,Cohn JN. Acute vasoconstrictor response to intravenous furosemidein patients with chronic congestive heart failure. Activation of theneurohumoral axis. Ann Intern Med 1985;103:1–6.

[19] Struyker Boudiw HA, Smits JF, Kleinjans JC, van Essen H. Hemo-dynamic actions of diuretic agents. Clin Exp Hypertens A1983;5:209-223.

[20] Johnston GD, Nicholls DP, Leahey WJ. The dose-response charac-teristics of the acute non-diuretic peripheral vascular effects offrusemide in normal subjects. Br J Clin Pharmacol 1984;18:75-81.

[21] Mukherjee SK, Katz MA, Michael UF, Ogden DA. Mechanisms ofhemodynamic actions of firrosemide: differentiation of vascular andrenal effects on blood pressure in functionally anephric hypertensivepatients. Am Heart J I981;1OI:313–318,

[22] Johnston GD, Hiatt WR, Nies AS, Payne NA, Murphy RC, GerberJG. Factors modifying the early nondiuretic vascular effects offurosemide in man, The possible role of renal prostaglandins. CircRes 1983;53:630–635.

[23] Gerkens JF, Smith AJ. Inhibition of vasoconstriction by frusemide inthe rat, Br J Pharmacol 1984;83:363–371.

[24] Gerkens JF, Armsworth SJ, Bbagwandeen SB, Smith AJ. Chemicalrenal medrdlectomy prevents frusemide-induced inhibition of sympa-thetic vasoconstriction in the in situ blood perfused rat mesentery. JHypertens 1987;5:273–275.

[25] Gerkens JF, Armsworth SJ, Dosen PJ, Smith AJ. Endothelium-de-pendent inhibition of sympathetic vasoconstriction by frusemideadministration to rats. Clin Exp Pharmacol Physiol 1988;15:449–455.

[26] Davis Jo, Freeman RH. Mechanisms regulating renin release. Phys-iol Rev 1976;56:1–56.

[27] Oates JA, Whorton AR, Gerkens JF, Branch RA, Hollifield JW,Frolich JC. The participation of prostaglandins in the control ofrenin release. Fed Proc 1979;38:72–74.

[28] Patrono C, Pugliese F. The involvement of arachidonic acidmetabolism in the control of renin release. J Endocrinol Invest1980;3:193-201.

[29] Francisco LL, Osborn JL, DiBona GF, Prostaglandin in renin releaseduring sodium deprivation. Am J Physiol 1982;243:F537–F542,

[30] Data JL Rane A, Gerkens J, Wilkinson GR, Nies AS, Branch RA.

[31

[32

The influence of indomethacin on the pharmacokinetics, diureticresponse and hemodynamics of fumsemide in the dog. J PharmacolExp Ther 1978;206:431–438.Kopp U, Aurell M, Sjolander M, Ablad B. The role of prosta-glandins in the alpha- and beta-adrenoceptor mediated renin releaseresponse to graded renal nerve stimulation, Pfliigers Arch1981;391:1–8.Gerber JG, Nies AS, Olsen RD. Control of canine renin release:macrrla densa requires prostaglandin synthesis. J Physiol1981;319:419–429.

[33] He XR, Greenberg SG, Briggs JP, Schnermann J. Effects offurosemide and verapamil on the NaCl dependency of’ maculadensa-mediated renin secretion. Hypertension 1995;26: 137–142,

[34] Barden AE, Mahoney DP, Tunney AM, Vandongcn R. Frusemidereleases renin in the rat kidney when prostacyclin synthesis issuppressed. Br J Pharmacol 1984;82:493–499.

[35] Packer M. Interaction of pmstaglandins and angiotensin 11 in themodulation of renal function in congestive heart failure. Circulation1988;77:164–173.

[36] Johnston GD, Nicholls DP, Leahey WJ, Finch MB. The effects ofcaptopril on the acute vascular responses to frusemide in man, ClinSci 1983;65:359-363.

[37] Collier JG, Robinson BF. Comparison of effects of locally infusedangiotensin I and II on hand veins and forearm arteries in man:evidence for converting enzyme activity in limb vessels. C’lin SciMol Med 1974;47:189–192.

[38] Passmore AP, Whitebead EM, Johnston GD. Comparison of theacute renal and peripheral vascular responses to frusemide andbumetanide at low and high dose. Br J Clin Pharmacol1989;27:305–312.

[39] Johnston GD, Nicholls DP, Kondowe GB, Finch MB. Comparisonof the acute vascular effects of frusemide and bumetanide. Br J ClinPharmacol 1986;21:359–364.

[40] Gerber JG, Nies AS. Furosemide-induced vasodilation: importanceof the state of hydration and filtration. Kidney Int i980;18:454–459.

[41] Williamson HE, Bourland WA, Marchand GR. Inhibition offurosemide induced increase in renal blood flow by indomethacin.Proc Soc Exp Biol Med 1975;148:164–167.

[42] Patrono C, Dunn MJ. The clinical significance of inhibition of renalprostaglandin synthesis. Kidney Int 1987;32:1–12.

[43] Wiemer G, Fink E, Linz W, Hropot M, Scholkens BE, Wohlfart P.Furosemide enhances the release of endothelial kinins, nitric oxideand prostacyclin. J Pharmacol Exp Ther 1994;271:1611–1615.

[44] Gerber JG, Hubbard WC, Branch RA, Nies AS. The lack of aneffect of furosemide on uterine prostaglandin metabolism in vivo.Prostaglandins 1978;15:663–670.

[45] Bailie MD, Cmsslan K, Hook JB. Natriuretic effect of’firmsemideafter inhibition of prostaglandin synthetase. J Pharmacol Exp Ther1976;199:469–476.

[46] Horrobin DF, Manku MS, Mtabaji JP. Vascular actions of frusemideand bumetanide on the rat superior mesenteric vascular bed: interac-tions with prostaglandins. Clin Sci Mol Med Suppl 1976;3:257s–258s.

[47] Lundergan CF, Fitzpatrick TM, Rose JC, Ramwell PW, Kot PA.Effect of cyclooxygenase inhibition on the pulmonary vasodilatorresponse to furosemide. J Pharmacol Exp Ther 1988;246:102–106.

Dow



ovember 2021

T.P.J. Dormans et al. /Cardioua.rcular Research 32 (1996) 988-997 997

[48] Mackay IG, MuirAL,WatsonML.Contributionof prostaglandinsto thesystemicandrenalvascularresponseto frusemide in normalman. Br J Clin Pharmacol 1984;17:513–519.

[49] Mtabaji JP, Manku MS, Horrobin DF. Vascukiraetionsof furosemideand bumetanide on the rat superior mesenteric vascular bed: interac-tions with prolactin and prostaglandins. Can J Physiol Pharmacol1976;54:357-366.

[50] Scherer B, Weber PC. Time-dependent changes in prostaglandinexcretion in response to frrrsemide in man. Clin Sci 1979;56:77-81.

[51] Sullivan JM, Patrick DR. Release of prostaglandin 12-like activityfrom the rat aorta: effect of captopril, furosemide, and sodium.Prostaglandins 1981;22:575–585.

[52] Katayama S, Attallah AA, Stahl RA, Bloch DL, Lee JB. Mechanismof furosemide-induced natriuresis by direct stimulation of renalprostaglandin E2. Am J Physiol 1984;247:F555-F561.

[53] Yared A, Kon V, Ichikawa I. Mechanism of preservation ofglomerular perfusion and filtration during acute extracellular fluidvolume depletion. Importance of intrarenal vasopressin–prosta-glandin interaction for protecting kidneys from constrictor action ofvasopressin. J Clin Invest 1985;75:1477–1487.

[54] Patrono C, Pugliese F, Ciabattoni G, et af. Evidence for a directstimulator effect of prostacyclin on renin release in man. J ClinInvest 1982;69:231-239.

[55] Benjamin N, Cafver A, Collier J, Robinson B, Vallance P, Webb D.Measuring forearm blood flow and interpreting the responses todrugs and mediators. Hypertension 1995;25:918-923.

[56] Aellig WH. Clinical pharmacology, physiology and pathophysiologyof supetilcial veins—1. Br J Clin Pharmacol 1994;38:181–196.

[57] Aellig WH. Clinical pharmacology, physiology and pathophysiologyof supefilcial veins—2. Br J Clin Pharrnacol 1994;38:289-305.

Dow



ovember 2021

vascular effects of loop diuretics - cardiovascular research

Documents