Acute Preload Effects of Furosemide
Peter A. Kraus, M.B., B.Ch.;* jeffrey Lipman, M.B. , B.Ch.;* and
Pieter J Becker, Ph.D.t
Acute preload effects (as re8ected by the pulmonary
capillary wedge pressure [PCWP]) of an intravenous furo-
semide bolus were studied in 33 patients. In those patients
receiving no vasoactive drug and in those receiving predom-
inantly preload reducing agents, there was an initial rise in
PCWP up until 15 minutes followed by a diuresis-induced
fall in PCWP below baseline levels at 1 h. Patients who
were receiving preload and significant afterload reduction
showed an immediate drop in PCWP which was sustained.
I ntravenous furosemide is traditionally administered
to patients in order to relieve pulmonary vascular
congestion. Initially, this was thought to be due to
rapid diuresis induced by the drug, thereby reducing
plasma volume and hence the pulmonary capillary
pressure.
Biagi and Bapat 1 in 1967 were the first to suggest
the possibility of extrarenal mechanisms, as they
observed relief of pulmonary edema in a patient before
the onset of diuresis. In 1969, Bhatia et al 2 studied
seven patients suffering from high altitude pulmonary
edema and similarly concluded that there was a
significant reduction in pulmonary blood volume
before the onset of diuresis. In 1973, Dikshit et al 3
studied renal and extrarenal hemodynamic effects of
furosemide in patients in left ventricular failure (LVF)
following acute myocardial infarction (MI). They doc-
umented that prior to a significant diuresis beginning
at 15 minutes post-administration of intravenous fu-
rosemide there was a fall in left ventricular end-
diastolic pressure as reflected by the pulmonary
capillary wedge pressure (PCWP). Their results sug-
gested that there was an earlier increase in venous
capacitance either due to a direct pulmonary vascular
vasodilatory effect or a peripheral venous pooling
effect.
Although this work had detractors from early on
(Hesse et al 4 ) it remained the standard reference for
the effects of intravenous furosemide in relieving
pulmonary vascular congestion. However, recently
Francis et al5 reported that in chronic congestive
cardiac failure bolus doses of intravenous furosemide
raised the PCWP initially, followed by a diuresis-
*Department of Anesthesia and Intensive Care Unit, Baragwanath
Hospital and University of the Witwatersrand, Johannesburg,
South Africa.
tinstitute for Biostatistics, South African Medical Research Council.
Manuscript received September 22; revision accepted January 2,
1990.
124
This trend is independent of underlying pathology or dose
of furosemide used. It is postulated that furosemide causes
an early deleterious release of endogenous vasoconstrictors
which may be blocked by combined preload and afterload
reduction. (Chest 1990; 98:124-28)
PCWP =pulmonary capillary wedge pressure; PAC= pul-
monary artery catheter; TNT= nitroglycerine
induced drop in PCWP. Nishimura and Kanbe6 also
investigated the effects of intravenous furosemide
following acute MI and showed a similar trend.
With the above controversy in mind, we decided to
study a cross-section of our intensive care unit (ICU)
population in order to determine the acute PCWP
effects of an intravenous bolus dose of furosemide .
Furthermore, since the work by Francis et al5 had
suggested that there may in fact be a deleterious
neurohumoral vasoconstrictor response to furosemide
usage, we studied the above effects with and without
the concomitant use of vasodilator drugs.
MATERIALS AND METHODS
All patients in our ICU during a 12-month period who had had a
pulmonary artery catheter (PAC) already in situ and were receiving
intravenous furosemide were eligible for the study. The PAC had to
be in situ for at least 6 h, there was to be no significant variation in
PCWP during this period, and no furosemide was to have been
given during this pretrial period . By the nature of our unit (combined
medical and surgical ICU), the majority of the patients studied
were in acute LVF following a MI, chronic LVF due to cardiomy-
opathy (CMO), or had the adult respiratory distress syndrome
(ARDS). Patients were divided into three groups (with a minimum
of ten patients in each group) according to whether vasodilator
drugs were being used at the time of the study or not. Patients were
eligible for study on more than one occasion. If patients were
receiving vasodilator drugs, the dosage of these agents was not
adjusted during the trial period.
Group 1 consisted of patients receiving no vasodilatory drugs at
all. Group 2 consisted of patients on therapy with predominant
preload reduction vasodilator. They received either low-dose oral
isosorbide dinitrate or low dose intravenous nitroglycerine (TNT)
both of which cause predominant venodilatation and hence preload
reduction.7-V Group 3 consisted of patients on preload and significant
after load reducing vasodilator therapy ("mixed" vasodilator therapy).
These patients received high dose TNT" or oral captopril in order
to achieve the "mixed" vasodilatation.
The PCWP was measured from the transduced pressure trace
on a graphic screen at the end of expiration"' just prior to
administration of intravenous furosemide . This was recorded as
time 0. If the PCWP was 20 mm Hg or less, 20 mg furosemide was
given. If the PCWP was 21 to 25 mm Hg, 40 mg of furosemide was
Acute Preload Elfecls of Furosemide (Kraus, Upman, Becker)
 Table 1- Furosemide Only

Dose
PT Age , Sex, Furosemide End-Expiratory PCWP in mm Hg (at 5 min intervals)
No, Diagnosis (mg) 5' w ~ - ~ ~ ~ • w ~ 55' 60'

1,53,M ARDS 20 14 15 14 15 12 12 12 12 12 12 12 12 12
2,56,F ARDS 20 18 16 16 16 16 15 15 15 16 14 15 12 13
3,51,M ARDS 20 20 20 20 22 22 20 20 22 20 20 20 20 20
4,37,F MI 20 18 20 20 21 18 20 20 20 20 21 18 17 15
5,55,F PE 20 20 20 20 18 19 18 18 17 15 15 16 15 15
6,19,F ARDS 20 15 16 16 15 14 12 10 10 9 10 11 12 12
7,19,F ARDS 20 14 13 16 15 16 15 15 15 15 15 14 14 15
8,49,M ARDS 20 19 19 20 18 17 18 18 17 16 16 16 17 16
9,19,F ARDS 20 16 16 16 16 15 15 16 15 15 15 15 15 15
10,67,M ARDS 40 21 20 20 20 20 20 19 19 20 20 20 20 20
11,23,F ARDS 20 18 17 17 17 16 15 15 15 15 15 15 15 15
12,64,M CMO 20 20 22 24 22 22 24 20 22 20 24 22 25 24

Table 2-Furosemide Plus Preload Reduction

Dose
PT Age, Sex Furosemide Vasoactive End-Expiratory End PCWP mm Hg (at 5 min intervals)
No, Diagnosis (mg) Drug 0' 5' 10' 15' 20' 25' 30' 35' 40' 45' 50' 55' 60'

1,48,M MI 40 23 24 24 24 24 22 20 20 19 20 20 20 18
2,53,M MI 40 TNT 21 20 20 19 18 18 20 19 15 14 16 16 16
3,62,F MI 40 TNT 23 24 25 25 25 25 27 25 23 22 22 22 22
4,59,M MI 20 I 15 15 15 14 14 12 12 10 10 10 10 10 12
5,59,M MI 20 TNT 18 20 20 20 20 19 15 15 15 16 15 14 17
6,59,M MI 20 TNT 17 19 18 17 17 17 20 20 20 21 22 20 20
7,59,M MI 80 TNT 30 30 28 28 28 27 27 26 26 26 26 25 25
8,59,M MI 20 I 19 18 19 17 16 16 15 14 12 15 15 13 14
9,28,M CMO 80 TNT 30 30 32 30 29 28 28 28 27 32 30 30 27
10,28,M CMO 80 TNT 27 30 29 29 32 30 29 29 30 30 30 30 29

given. If the PCWP was >25 mm Hg, 80 mg of furosemide was found by doing pairwise comparisons at the Bonferroni adjusted
given. Following the administration of furosemide , the PCWP was level of significance (0.0513=0.017), again taking into account
measured every 5 minutes for the next hour by the same observer. whether the variances in the groups were equal or not.

Statistical Methods RESULTS

At each time and at the 0.05 level of significance, the groups There were 12 patients in group 1 (furosemide only),
were compared with respect to the mean change from baseline in a ten patients in group 2 (furosemide plus preload
one-way analysis of variance. At those times where, according
reduction) and 11 patients in group 3 (furosemide plus
Levene's test , the variances in the groups were not equal, the Welch
test was employed to compare the groups with respect to the mean preload and afterload reduction). Group 2 patients
change from baseline. Specific differences between the groups were received up to a 100 J.Lg/min intravenous TNT or up
Table 3-Furosemide Plus Preload and Afterload Reduction

Dose
PT Age , Sex Furosemide Vasoactive End-Expiratory PCWP mm Hg (at 5 min intervals)
No, Diagnosis (mg) Drug 0' 5' 10' 15' 20' 25' 30' 35' 40' 45' 50' 55' 60'

1,61,F MI 40 I, C 21 20 22 20 20 20 20 20 21 27 20 20 20
2,53,F MI 40 TNT 25 25 25 24 22 20 25 24 22 22 22 20 19
3,58,M MI 40 TNT 23 22 18 18 19 18 17 18 18 19 19 20 20
4,46,M CMO 20 c 20 20 18 18 18 18 18 17 17 17 19 17 17
5,78,F MI 40 TNT 21 24 21 17 14 14 15 14 20 17 16 16 15
6,18,F CMO 40 c 25 20 18 16 17 17 15 15 15 15 16 17 16
7,18,F CMO 40 c 25 25 22 21 18 19 20 20 20 18 22 23 20
8,66,M MI 40 TNT 22 18 15 20 20 20 18 18 18 20 20 16 18
9 ,59,M MI 80 TNT 27 27 27 27 27 26 27 27 27 26 27 27 25
10,78,F MI 40 TNT 25 27 25 26 25 22 20 19 22 19 15 18 19
11,50,M MI 40 TNT 22 20 17 17 15 12 12 10 10 10 12 12 10

ARDS =adult respiratory distress syndrome; MI =myocardial infarction; PE =pulmonary embolism; CMO =cardiomyopathy; I= isosorbide
dinitrate; TNT= nitroglycerine; C =captopril

CHEST I 98 I 1 I JULY. 1990 125

 Table 4-Mean Change in PCWP in mm Hg (5 Min Intervals)

0' 5' 10' 15' 20' 25' 30' 35' 40' 45' 50' 55 ' 60'
Croup 1 0 0 +0.5 +0.16 -0.5 -0.75 -1.25 -1.16 -1.66 -1.33 -1.58 -1.75 -1.75
Croup 2 0 +0.7 +0.7 0 0 -0.9 -1.0 -1.7 -2.6 -1.6 -1.5 -2.3 -2.2
Croup3 0 -0.72 -2.45 -2.9 -3.72 -4 .54 - 4.45 -4 .9 -4.18 - 4. 18 -4 .27 -4 .36 -4.81

to 20 mg isosorbide dinitrate every 8 h. Group 3
patients received doses which were more than the
above. With reference to Tables 1, 2 and 3, it should
be noted that patients in group 1 were younger than
those in groups 2 and 3 (average age 43 years vs 51
and 53 years). The diagnosis was usually ARDS in
group 1 (MI and CMO in groups 2 and 3). The initial
PCWP was lower than those in groups 2 and 3 (average
18 mm Hg vs 22 and 23 mm Hg); hence, the average
dose of furosemide was less (20 mg vs 40 mg). Patients
in groups 2 and 3 were similar in age, diagnosis, initial
PCWP and dose of furosemide . Sex distribution in all
groups was similar, except for group 3 who had a
preponderance of men. Table 4 shows the mean change
in PCWP from baseline in the various groups.
The most notable feature on comparing these three
groups is that the patients in group 3 (furosemide plus
preload and afterload reduction) generally showed an
immediate fall in PCWP following furosemide admin-
istration. In comparison, groups 1 and 2 patients
initially showed an increase in PCWP up until 15
minutes, thereafter showing a drop but not to the
same degree as group 3 patients. From the appropriate
one-way analysis of variance, the three groups were
found to be significantly different (p<0.05) with re-
spect to the change from baseline for the times 5 to
35 minutes. In particular, from the pairwise compari-
sons done at the Bonferroni adjusted level of signifi-
cance groups 1 and 2 were not significantly different,
but both changed significantly less than group 3. This
trend is depicted in Figure 1 where the mean change
in PCWP from the initial PCWP is plotted according
to time post administration offurosemide.
DISCUSSION
Nitrates are often used to decrease preload. Unfor-
tunately, for a clinical study, TNT and isosorbide
dinitrate are not pure preload reducing agents and
especially in higher doses reduce afterload too. 7· 9 We
divided the patients receiving nitrates into those with
predominant preload effects (low dose, group 2) from
those with both preload and afterload effects (high

•·----.•
3 o-·-·-·0 Group 1 (Furosemide only)
Group 2 (Furosemide + Preload)
2 Group 3 (Furosemide+ Preload + Afte
......
Cl
I
E
E
Q)
c 0
Q)
Cl)
ctl
.0 -1
E
0
....
-2
Q)
Cl
c
ctl
..c -3
()

-4

-5

0 5 10 15 20 25 30 35 40 45 50 55 60
Time (mins)
FIGURE I. Mean change from baseline in pulmonary capillary wedge pressure over first hour. Croup 1:
furosemide only; group 2: furosemide plus preload reduction; group 3: furosemide plus preload plus
afterload reduction. Differences between groups 1 and 2 vs group 3 are statistically significant (p < O.OS)
between 5 and 35 minutes (see text for more details).

126 Acute Preload Effects of Furosemide (Kraus, Lipman, Becker)

dose, group 3). This is a technically artifactual differ-
entiation, but does have some clinical relevance.
The doses of furosemide given were compatible
with those used in most of the controversial
literature3 ·s·6 studying its effects on hemodynamics.
The design of the study was to keep as close to the
normal clinical situation as possible . Larger doses have
been used, but like Kiely et al 11 we prefer to use lower
repeated doses to achieve a desired effect rather than
running the risk of a precipitous fall in cardiac output
and systemic pressure.
Although group 1 patients in general had a different
underlying pathology than groups 2 and 3, the change
in PCWP was virtually the same as group 2 patients
(Fig 1). Following administration of furosemide they
showed an actual increase in PCWP up until 10
minutes, only falling below baseline levels after 15
minutes. Group 3 patients had an immediate drop in
PCWP which was sustained and of greater magnitude
than groups 1 and 2 (Fig 1). The causes of pulmonary
congestion in groups 2 and 3 patients were similar, yet
the response to furosemide of the PCWP differed
significantly. Hence, we feel that our results were not
influenced by the nature of the underlying pathology
but by the concomitant vasodilator therapy used. 3·M
Patients who received furosemide only showed an
initial increase in the PCWP up until approximately
15 minutes. This is at variance with the study by
Dikshit et al,3 yet is in agreement with the more
recent findings of Nishimura and Kanbe6 and Francis
et al.s Dikshit et al3 had shown an increase in venous
capacitance to account for an early drop in PCWP, but
these results were questioned by Hesse et al. •
Dikshit et al3 studied 20 acute MI patients with
LVF and showed that following intravenous furosemide
the PCWP dropped significantly within the first 15
minutes accompanied by a significant increase in calf
venous capacitance. Diuresis reached a peak at 30
minutes causing a further drop in PCWP. Hesse et al4
studied nine patients not in cardiac failure and showed
that, although there was a gradual drop in PCWP,
there was no change in forearm venous tone and
attributed this fall purely to diuresis. In addition, he
documented increased renin activity prior to this
diuresis. Francis et al5 documented a similar increase
in renin activity prior to diuresis in 15 patients in
chronic heart failure, as well as increased noradrena-
line and vasopressin levels. The resultant early vaso-
constriction was their explanation for the PCWP
actually going up in the first 20 minutes post-admin-
istration of furosemide . They felt that renin was
released earlier due to direct mechanisms rather than
later as a result of diuresis-induced plasma volume
depletion. This had also been suggested by the earlier
work of Hesse et al 4 and Rosenthal et al. 12 Francis et
als also queried the possibility of using angiotensin-
converting enzyme (ACE) inhibitors to block this
initial rise in PCWP. By giving "mixed" vasodilator
therapy (high dose nitrates, ACE inhibitors) we have
blocked this rise in PCWP. Nishimura and Kanbe6
studied 30 patients with acute MI and showed that
following administration of furosemide the PCWP
originally increased only falling below baseline after
15 minutes. Our study showed very similar results
(Fig 1, groups 1 and 2).
We have thus shown that by using "mixed" vasodi-
lators (pre- and afterload reducing agents) the initial,
possibly deleterious, effects of furosemide may be
eliminated. Conversely, furosemide-induced release
of endogenous vasoconstrictors would explain why
group 1 and 2 patients had an initial rise in PCWP.
The subsequent fall in PCWP in all groups can then
be attributed to the diuretic effect on plasma volume.
In addition, the greater overall drop in PCWP in those
who have afterload reduction has been documented
previously by Nelson et al.11
The fall in PCWP in all groups from approximately
15 minutes onward is universally accepted as being
due to the furosemide-induced diuresis. We therefore
did not see the need to document urine output over 5
min intervals during the trial period. In addition, most
of the patients were not catheterized, and hence, an
accurate time course for urine output could not be
determined. Nevertheless, the urine output at the end
of the hour showed a marked increase from the
previous hourly urine outputs (not shown here) and
was dose-dependent, as previously shown.6
The effects of furosemide on hemodynamics and
the relief of pulmonary congestion remains contentious
according to the world literature. The dose of furosem-
ide plus the duration of the pulmonary congestion
(according to underlying pathology) are often cited as
the reasons for these discrepant results. We feel that
our study cuts across these differences and shows that
intravenous furosemide without concomitant "mixed"
vasodilation increases the PCWP initially probably
due to increased endogenous vasoconstricting hor-
mones. Thereafter, there is a drop in PCWP due to a
diuresis-induced drop in plasma volume. This trend
is independent of underlying pathology or dose of
furosemide used .
In summary, our study shows that furosem ide, when
given as a bolus dose , does not immediately decrease
pulmonary pressures. In fact, unless "mixed" vasodi-
lating agents are given concomitantly, there is a rise in
PCWP. This would be in keeping with results of
Francis et al5 and Nishimura and Kanbe. 6 Only when
pre- and afterload reduction is used is there an
immediate fall in PCWP. One of the reasons we
undertook this study was that occasionally a patient
in pulmonary edema, given furosemide, in our opinion
would deteriorate clinically prior to improving. AI-
CHEST I 98 I 1 I JULY, 1990 127
though not specifically addressed in the protocol, this
could be explained by an early increase in PCWP
after administration of furosemide . The increased
PCWP could lead to increased interstitial water, 13 and
hence, clinical deterioration. However, whether the
increase we documented is clinically significant in
relation to pulmonary edema, we cannot answer at
this stage . If the clinician wants to prevent this rise in
PCWP and cause a fall in venous pressures, before
the onset of furosemide-induced diuresis, we suggest
concomitant use of a "mixed" vasodilator, eg, high
dose nitrates or ACE inhibitors.
ACKNOWLEDGMENT; The authors wish to acknowledge Sr. J.
Scribante for the pilot study of this work and Hoechst Pharmaceu-
ticals for their financial support.
REFERENCES
Biagi RW, Bapat BN. Frusemide in acute pulmonary oedema.
Lancet I967; I:849
2 Bhatia ML, Singh I, Manchanda SC, Khanna PK, Roy SB.
Effect of frusemide on pulmonary blood volume. Br Med J
I969; 2:55I-52
3 Dikshit K, Vyden JK, Forrester JS, Chatterjee K, Prakash R,
Swan HJC. Renal and extrarenal hemodynamic effects of furo-
semide in congestive heart failure after acute myocardial infarc-
tion. NEngl J Med I973; 288:1087-90
4 Hesse B, Nielsen I, Lund-Jacobsen H. The early effects of
intravenous frusemide on central haemodynamics, venous tone
128
and plasma renin activity. Clin Sci Mol Med 1975; 49:551-55
5 Francis GS, Siegel RM, Goldsmith SR, Olivari MT, Levine TB,
Cohn JN. Vasoconstrictor response to intravenous furosemide
in patients with chronic congestive heart failure-Activation of
the neurohumoral axis. Ann Intern Med 1985; 103:I-6
6 Nishimura N, Kanbe N. The renal and hemodynamic effects of
furosemide in acute myocardial infarction. Crit Care Med 1981;
9:829-32
7 Rabinowitz B, Tamari I, Elazar E , Neufeld HN. Intravenous
isosorbide dinitrate in patients with refractory pump failure and
acute myocardial infarction. Circulation 1982; 65:771-78
8 Nelson GIC, Silke B, Forsyth DR, Venna SP, Hussain M, lllylor
SH. Hemodynamic comparison of primary venous or arteriolar
dilatation and the subsequent effect of furosemide in left
ventricular failure after acute myocardial infarction. Am J
Cardiol1983; 52:1036-40
9 Flaherty JT, Come PC, Baird MG, Rouleau J, Taylor DR,
Weisfeldt ML, et al. Effects of intravenous nitroglycerin on left
ventricular function and ST segment changes in acute myocardial
infarction. Br Heart J I976; 38:612-21
IO Berryhill RE, Benumof JL, Rauscher LA. Pulmonary vascular
pressure reading at the end of exhalation. Anesthesiology 1978;
49:365-68
II Kiely J, Kelly DT, lllylor DR, Pitt B. The role of furosemide in
the treatment of left ventricular dysfunction associated with
acute myocardial infarction. Circulation I973; 48:581-87
12 Rosenthal J, Boucher R, Nowaczynski W, Genest J. Acute changes
in plasma volume, renin activity, and free aldosterone levels in
healthy subjects following fursemide administration. Can J Physiol
Phannacol1968; 46:85-91
I3 Staub NC. The pathogenesis of pulmonary edema. Prog Cardio-
vasc Dis I980; 23:53-80
Acute Pnlload Effects of Furosemide (Kraus, Upman, Becker)