Intermittent Administration of Furosemide vs Continuous Infusion Preceded by a Loading Dose for Congestive Heart Failure

Intermittent Administration of Furosemide vs Continuous Infusion Preceded by a Loading Dose for Congestive Heart Failure* Meir lAhav, M.D.; Arie Regev, M.D.; Pia Ra'anani, M.D.;

and Emanuel Theodor, M.D.

Several reports have suggested that continuous intravenous administration of loop diuretics may be superior to intermittent administration. We performed a prospective randomized crossover study comparing intermittent intravenous administration (IA) of furosemide with continuous infusion following a single loading dose (LDCI) in nine patients with severe congestive heart failure. At the time of hospital admission, patients were randomly assigned to one of two treatment groups. One group (four patients) received an IV bolus injection of furosemide fOllowed immediately by a continuous infusion for 48 h. The second group (6ve patients) was treated with three IV bolus injections a day for 48 h. Total doses of furosemide were equivalent in

the two groups. After 48 b, each patient was crossed over to the other method and treated for an additional 48 h. LDCI produced significantly greater diuresis and natriuresis than IA (total urine output increased by 12 to 26 percent, total sodium excretion increased by 11 to 33 percent) (p
Furosemide is a potent loop diuretic that is widely employed in the treatment of congestive heart failure (CHF). When administered intravenously (IV), it induces a prompt and vigorous diuresis with a wide range dose response curve. 1 In severe heart failure, relatively high doses of furosemide are used. These doses are often associated with significant side effects, including hypovolemia, electrolyte disturbances, hyperuricemia, and metabolic alkalosis.2.3 Furosemide is usually administered intermittently by IV bolus injections. This mode of administration may lead to marked fluctuations in intravascular volume and to high peak serum levels of furosemide, increasing its toxicity." It seems conceivable that continuous furosemide infusion may decrease the Huctuations in intravascular volume causing a relatively constant hourly urine output. s It may also prevent the accumulation of toxic levels of furosemide, thus causing fewer and less severe side effects. 6 Moreover, this method of administration would allow rapid termination of an undesirably vigorous diuresis or a side effect. In patients refractory to conventional doses offurosemide, continuous IV administration may allow gradual increase of infusion rate until the desirable hourly diuresis occurs. 6 •7 Several reports have suggested that continuous IV

administration of loop diuretics may also lead to increased diuretic and natriuretic effects compared with intermittent schedule.5-9 This concept received theoretical support from several authors l ()"12 but only two prospective controlled clinical studies addressed this issue directly.5.13 One of these studies demonstrated no significant difference in the diuretic and natriuretic effects of the two modes of administration in patients after open heart surger)'.s The other study demonstrated superior efficacy of the continuous IV administration in patients with chronic renal failure. 13 However, the patients examined in this study did not have CHF. Since the dose response relationship of furosemide is altered considerably in CHF,12 the results of this study cannot be extrapolated to patients with CHF. Furthermore, in both of these studies, furosemide was administered for relatively short periods. It has been demonstrated that continuous administration of furosemide causes a gradual increase in urine output that may reach a peak only several hours after the initiation of infusion. S This lag period may be critically important in patients with severe CHF. Based on these observations, we postulated that continuous infusion of furosemide preceded by an IV loading dose may induce increased diuresis with fewer side effects compared with the conventional intermittent mode of administration in patients with CHF.

*From the Department of Internal Medicine E, Beilinson Medical Center, Petali TIqva and Tel Aviv University Sackler School of Medicine, Israel. Manuscript received October 9; revision accepted February 25.

METHODS

=

IA intermittent administration; LDCI = continuous infusion

following loading dose

lbtients All the patients included in this study were admitted to the CHEST I 102 I 3 I SEPTEMBER, 1992

725

Table I-Patient Dow· Patient/Sexl Age, yr l/MI69

4IMnl 5/F18O

8IF/68

9IMns

Diagnosis IHD, Lt heart failure, SIP ant and inf wall MIs IHD, Lt heart failure, SIP ant lat wall MI RHD, Rt and Lt heart failure, mitral valve disease, chronic atrial 6brillation IUD, SIP ant and inf MIs, Lt heart failure COPD,cor pulmonale, chronic atrial fibrillation, hypertension, Lt and Rt heart failure IUD, hypertension, SIP ant wall MI, Lt heart failure IUD, SIP ant wall MI, acute infwall MI, Lt heart failure RHD, chronic atrial fibrillation, insufficiency of prosthetic mitral valve due to SBE, Lt heart failure IUD, SIP ant wall MI, Lt heart failure

Medications in Addition to Furosemide Nitrates, amiodarone

Nitrates, captopril, aspirin Digoxin, warfarin

Captopril, thyroxine, allopurinol Digoxin, nitrates, aspirin, glibenclamide

Captopril, nifedipine, quinidine, pentoxifylline Nitrates, nifedipine

Digoxin, warfarin, cefuroxime, amikacin

Nitrates, captopril

*IHD = ischemic heart disease; RUO = rheumatic heart disease; SIP =state post; Lt =left; Rt = right; MI =myocardial infarction; COPD =chronic obstructive pulmonary disease; SBE = subacute bacterial endocarditis; ant =anterior; inf =inferior; lat =lateral. hospital with class 3-4 CHF (New York Heart Association classification) that was refractory to conventional oral therapy (ie, combinations of oral diuretics, nitrates, angiotensin-converting enzyme [ACE] inhibitors, and digoxin). Relevant clinical data are presented in Table 1. Patients were excluded from the study for the follOwing reasons: serum creatinine levels of >2.0 mgldl, serum sodium concentration < 130 mEqIL, liver disease, history of allergic reaction to furosemide, and additional medical therapy known to in8uence urine output, eg, aminophylline, dopamine, or thiazide diuretics. At the time of hospital admission, patients were randomly assigned to one of two treatment groups. One group received IV bolus injection of 30 to 40 mg of furosemide as a loading dose followed immediately by a continuous infusion of 2.5 to 3.3 mglh (60 to 80 mwday) for 48 h. The second group was treated with three IV bolus injections of 30 to 40 mg every 8 h for 48 h. Forty-eight hours after initiation of furosemide therapy, each patient was crossed over to the other treatment group and was treated for an additional 48 h. The total doses of furosemide were equivalent in the two modes ofadministration. Switching from the intermittent to the continuous mode of administration was done 8 h after the last IV bolus. Before crossing over from continuous infusion to intermittent administration, the patients were allowed a washout period of 3 h, during which they were not given furosemide. Each patienfs regular medications (ie, nitrates, ACE inhibitors, or digoxin) were held 728

constant throughout the study period. None of the patients received mechanical ventilation during the study. All details of this study were approved by the institutional clinical research committee and fully informed consent was obtained from each patient. Six of the nine patients included in the study had indwelling urinary catheters throughout the study period. In these patients, the urine output was measured every hour and urine samples were collected every 2 h for sodium and potassium determination. When an indwelling catheter was not inserted, urine output was measured and samples were collected after each urination. Blood samples were drawn at hospital admission, then once every 6 h for 12 h and every 12 h thereafter. Each sample was analyzed for sodium, potassium, chloride, serum urea nitrogen, and creatinine levels. The pH and bicarbonate were determined in venous blood every 12 h. Uric acid and calcium levels were determined every 24 h. Total 8uid intake was maintained at 22 ml/kg/24 h in all the patients. A low-sodium diet (daily sodium chloride intake of 2 g) was maintained throughout the study. The difference between the two methods of furosemide administration was assessed by net cumulative urine and sodium excretion. Statistical analysis of paired data was made using the Wilcoxon test for paired data. RESULTS

Furosemide was administered at two possible rates; six patients were treated with 120 mw24 h. This total dose was administered either as three bolus injections of 40 mg every 8 h or as a bolus of 40 mg followed by a continuous drip of 200 mg/48 h. Three patients received furosemide at a rate of 90 mg/24 h, either as three bolus injections of 30 mg every 8 h or as a bolus of 30 mg followed by a continuous drip of 150 mg/48

h.

Intermittent Administration (IA) Urine output peaked within 1 to 2 h after each bolus of furosemide and declined progressively thereafter, reaching baseline levels in 3 to 4 h (Fig 1). Mean hourly urine output ranged from 35± 18 m1lh (usually 4 to 8 h after furosemide injection) to 180 ± 62 mllh (1 to 2 h after injection). Hourly variations in urine output reached ± 120 ml during the first 3 h after bolus injections. Total urine output in 24 h ranged from 2,685 ml to 6,365 ml (mean, 3,790 ml) (Fig 2). Nearly 60 percent (2,230 ± 698 ml) of the mean total urine output of 48 h was excreted within the first 2-h periods after each bolus. Changes in sodium excretion during IA were similar to those of urine output (Fig 3). Total sodium excretion in 48 h ranged from 115 mEq to 547 mEq (Fig 4). Two patients were hyponatremic (130 mEq/L 8.0 mg/dl), hypochloremic alkalosis (pH >7.45), or Intermittent va Continuous Furosemide in CHF (Lahav at aI)

,, ~\

\

o

+

)y'

~M

..0

••

+

• • ~~

m~~a.~~M~.~G~

+

+

+

IMMIX

1. Mean hourly urine output in six catheterized patients during intermittent furosemide administration (open circles) and continuous infusion with a loading dose (black circles). FIGURE

hypomagnesemia (serum Mg <1.9 mEq/L) were not recorded during IA. None ofthe less common systemic side effects (eg, tinnitus, deafness, gastrointestinal symptoms, skin rash) developed during the treatment. Continuous Infusion following lAJading Dose (LDCI)

As in the intermittent mode ofadministration, urine output peaked 1 to 2 h after IV furosemide loading dose (Fig 1). After the second hour, there was a gradual decrease in hourly urine volumes, but diuresis was

maintained above preinfusion levels throughout the 48-h period. Mean hourly urine output ranged from 41 ± 13 mV h (before the initiation of furosemide) to 170 ± 55 mV h (2 h after IV loading dose). Hourly variations in urine output were maximal during the first 2 h after the IV loading dose (135 mJlh) but did not exceed 25 mJlh thereafter. Total urine output in 48 h ranged from 3,215 ml to 7,365 ml (mean, 4,490 ml) (Fig 2).

l§j Continuous

Li' Intermittent pc 0.01

....

:I

ea-

:I

o c

•

-c

.. :I

Ii

~

1

2

3

4

5

Petients

FIGURE 2. Total 48-h urine output with intermittent furosemide administration (IA) vs continuous infusion preceded by a loading dose (LDCI).

CHEST I 102 I 3 I SEPTEMBER. 1992

727

Q

c:

.2

1\

I

: I

I

\

iI +.z

"

1\

,, ,, , ,

I

, ,

I I

\

I

,

,,

\

I

\

\ ;1>--6

\ \

~

\

\ \

,,

1\ I

\

! ,

,

I

, I

\

,

b~ ,

b.-~

I I

\

, ,

, I

''t>-~

10 12 14 18 •

t

20 22 24 28 28 30 32 34 38 38 40 42 44 48 48 Houra

t

t

t

bol... I.Y.

FIGURE 3. Mean hourly sodium excretion in six catheterized patients during intermittent furosemide administration (open circles) and continuous infusion with loading dose (black circles).

As in intermittent administration, changes in hourly sodium excretion during continuous furosemide infusion paralleled those of urine output. Total sodium excretion in 48 h ranged from 135 mEq to 677 mEq. There was no significant difference in patterns of response between patients who were receiving ACE inhibitors or nitrates and those who were not. Two patients were hyponatremic (130 mEq/L
en

~~

LiJ Intermm.nt

~

I

J,

g

pcQ.01

! Ii • E

.

serum creatinine concentration was noted in one during LDCI. Hypokalemia occurred in two patients and hypocalcemia occurred in one. Hypochloremic alkalosis, hyperuricemia, or hypomagnesemia were not recorded during LDCI and none of the less common side effects were recorded during treatment. Total urine output with LDCI was consistently higher (by 12 to 26 percent; mean, 18.5 percent) than with IA (p
..

.2

!

.~

1.

218

.7

M

1

2

FIGURE 4. Total 48-h sodium excretion with intermittent furosemide administration (IA) vs continuous infusion preceded by a loading dose (LDCI).

728

Intermittent vs Continuous Furosemide in CHF (Lahav et 81)

DISCUSSION

Loop diuretics cause their natriuretic effect predominantly by decreasing chloride reabsorption in the thick ascending loop of Henle. Like several other diuretics, loop diuretics must reach the tubular lumen to be effective}"·15 They reach the luminal compartment by being actively secreted from the blood into the urine at the organic acid transport pathway of the straight segment of the proximal tubule. lO Using probenecid as a means to alter the relationship between serum and urine concentrations of furosemide, Chennavasin et al l4 have confirmed previous animal data suggesting that urinary furosemide is the best correlate of response. The relationship between urinary furosemide excretion rate and response is characterized by a sigmoid shaped curve. I ",16 This curve is suppressed in many patients with CHF and may occasionally lose its sigmoidal configuration in these patients. 12.16 Pretreatment of subjects with probenecid has been demonstrated to increase the overall natriuretic response to furosemide. This difference was attributed to the changed time course of delivery of furosemide into urine. 17 A similar phenomenon has been demonstrated by Kaojarern et alii when comparing overall response to oral IV dosing of furosemide. BraterlO used the concept of sodium to furosemide excretion ratio to describe the efficiency of the diuretic. By applying this approach, he found that the amount of urinary furosemide with maximal efficiency (21.5 ~w'min) was considerably less than the amount of drug causing half maximal response (ED50) (69.8 fLw'min). Brater has also demonstrated that after oral furosemide administration, the amount of furosemide in urine more persistently approached that amount with maximal efficiency. Thus, overall natriuresis was greater relative to the total amount of drug delivered to the acitve site. A similar effect was caused by pretreatment with probenecid. Based on these data, it seems conceivable that during slow IV infusion, the amount of furosemide in urine may approach the amount with maximal efficiency for longer periods of time.

This concept may explain greater overall diuresis with continuous infusion compared with IV bolus injections. 12 Recent publications dealing with continuous IV administration of furosemide include mostly case reports and uncontrolled studies of which only a few deal with patients with eHF (Table 2). Krasna et al7 reported a prompt resolution ofoliguria by continuous furosemide administration in three cardiac surgical patients with acute postoperative renal failure that did not respond to bolus injections of furosemide. Amiel et al8 reported rapid relief of tense ascites by IV influsion offurosemide in four patients with abdominal malignant neoplasms. Gerland and van Meijel9 treated 35 patients with CHF who were refractory to conventional therapy with high doses of furosemide (250 to 4,000 mg/day) either orally or IV (as bolus injections or continuous infusion). They did not conduct a comparative study between the different modes of administration, but they pointed out that slow continuous IV infusion seemed the most effective and least toxic. Lawson et al6 used a continuous IV infusion of furosemide in ten patients with CHF who failed to respond to 120 mg of furosemide orall~ All patients obtained "satisfactory diuresis with infusion of 4 to 16 mglh. Lawson et al noted that lower plasma concentrations of furosemide produced considerably higher sodium excretion rates during continuous IV administration. Vermeulen and Chadha l8 approached the same issue through the enteral route, by comparing a slow-release oral formulation with the standard formulation in patients with CHF. They found that total diuresis and natriuresis were equal in the two groups while peaks of diuresis were seen only in patients receiving the standard formulation. Only two studies presented a prospective randomized comparison between intermittent and continuous IV administration of loop diuretics. In one study (Copeland et al5), a gentle sustained diuresis was achieved by continuous infusion of furosemide in patients after cardiac surgery, but there was no significant difference in total urine volume and total u

Table 2-ContinuouslV Loop Diuretics: Recent Publications Source Krasna et a17 Arniel et al~ Gerland and van Meijel9 Lawson et a16 Copeland et al s Rudy et a1 13

No. of Subjects 3 4 10 18 8

Preinfusionl

Infusion Rate

Diagnosis

Control Dosage

Furosemide Acute renal failure after cardiac surgery Malignant ascites Class IV congestive heart failure

40 mg x 31day IV 80-120 mglday orally

0.38-0.75 mglkWh 100 mw24 h 250-4,000 mwday

Congestive heart failure Postcardiac surgery

120 mwday orally 0.3 mJ.Vkg x 2/12 h IV

4-16 mWh O.05mWkWh

6 mJtX2I12 h

1 mg bolus, 0.918 mglh

Chronic renal failure

Bumetanide

CHEST I 102 I 3 I SEPTEMBER. 1992

729

sodium excretion compared with bolus furosemide administration. Copeland et al5 measured urine output for only 12 h while diuresis in the continuous infusion group did not reach a peak until the third hour after initiation of infusion. In a pilot study of continuous IV administration in patients with CHF, we have seen that increase in urine output may continue for as long as 5 to 8 h after the initiation of infusion (data not shown). This lag period may account for the lack of significant difference between the two modes of administration in the study of Copeland et a1. Moreover, this lag period may be too long for patients with severe CHF. In a randomized crossover study published shortly after the initial submission of this article for publication, Rudy et a1 13 demonstrated that continuous infusion of bumetanide preceded by a bolus loading dose resulted in a significantly greater net sodium excretion compared with an intermittent bolus injection in eight patients with chronic renal failure. By using a loading dose prior to initiation of furosemide infusion, we induced prompt diuresis and therefore could proceed with relatively slow furosemide infusion rate to achieve sustained diuresis. In addition to inducing prompt diuresis, the initial IV bolus of furosemide has the advantage of its extrarenal hemodynamic effects in patients with CHF.19 These effects, though still controversial,20·21 are probably less likely to occur with continuous, slow infusion of furosemide. 22 In our stud); intermittent furosemide administration (IA) produced a series of peaks in urine output and natriuresis (Fig 1 and 3). Each peak occurred within the first 2 h after the bolus injection and was followed by gradual decrease, corresponding to the serum half-life of furosemide, which averages about 50 min (range, 30 to 70 min).23 LOCI produced initial peak in diuresis and natriuresis followed also by a gradual decrease, but here urine output and sodium excretion were maintained at a constantly increased level throughout the administration of furosemide. LDCI produced diuresis and natriuresis that were significantly and consistently greater than with IA (Fig 2 and 4) regardless of initial mode of administration (IA or LOCI). Furthermore, in contrast to the large fluctuations in hourly urine output in the IA group (Fig 1), LOCI induced sustained diuresis with relatively little variation in urine output. In our study, there were no significant differences in side effects between the two modes of administration, but such differences may appear following administration of larger doses of furosemide or in larger groups. There was a mild and gradual decline in hourly diuresis and natriuresis throughout the 48-h follow-up 730

in both modes of administration, although it was not statistically significant. This finding has been described by several authors and is probably related to changes in sodium balance.4.5.10.13 We have demonstrated that LDCI of furosemide produces a significantly greater diuresis and natriuresis than IA. It seems, therefore, that LOCI has several advantages over IA in CHF. We believe LDCI is also superior to continuous infusion without a loading dose, since LOCI may combine the advantages of continuous infusion with those of immediate IV bolus in patients with severe CHF. Because oftechnical limitations, we could not measure the serum and urinary concentrations of furosemide in our patients. Results of such measurements in future studies can conceivably provide additional data regarding the mechanisms underlying the differential responses between LDCI and IA. We also believe that further investigation of the renal and extrarenal effects of this method offurosemide administration is warranted. ACKNOWLEDGMENT: We thank the nursing staff of our department for providing help in this study.

REFERENCES 1 Benet 12. Pharmacokinetics/pharmacodynamics of furosemide in man: a review J Pharmaookinet Biopharm 1979; 7:1-27 2 Plumb VJ, James TN. Clinical hazards of powerful diuretics: furosemide and ethacrynic acid. Mod Concepts Cardiovasc Dis 1978; 47:91-4 3 Lowe J, Gray J, Henry DA, Lawson DH. Adverse reactions to furosemide in hospital inpatients. BMJ 1979; 2:360-62 4 Branck RA, Roberts CJC. Homeida M, Levine D. Determinants of response to furosemide in normal subjects. Br J Clin PharmacoI1977; 4:121-27 5 Copeland JG, Campbell D~ Plachetka JR, Salmon N~ Larson DF. Diuresis with continuous infusion of furosemide after cardiac surgery. Am J Surg 1983; 146:796 6 Lawson DH, Gray JMB, Henry DA, Ttlstone WJ. Continuous infusion of furosemide in refractory oedema. BMJ 1978; 2:476 7 Krasna JM, Scott GE, Scholz PM, Spotnitz AJ, Mackenzie JW: Penn F. Postoperative enhancement of urinary output in patients with acute renal failure using continuous furosemide the~ Chest 1986; 89:295 8 Amiel FA, Blackburn AN, Rubens RD. I.~ infusion offurosemide as treatment for ascites in malignant disease. BMJ 1984;

283:1041 9 Gerland PGG, van Meijel JJM. High dose furosemide in the treatment of refractory congestive heart failure. Arch Intern Moo 1988; 148:286 10 Brater DC. Pharmacodynamic considerations in the use of diuretics. Ann Rev Pbarmacol Toxicoll983; 23:45-62 11 Kaojarern S, Day B, Brater DC. The time course of delivery of furosemide into urine is an independent determinant of overall response. Kidney Int 1982; 22:69-74 12 Brater DC. Resistance to loop diuretics: why it happens and what to do about it. Drugs 1985; 30:427-43 13 Rudy D~ Voelker JR, Greene PK, Esparza FA, Brater DC. Loop diuretics for chronic renal insufficiency: a continuous infusion is more efficacious than bolus therapy. Ann Intern Moo 1991; 115:360-66 14 Chennavasin ~ SeiweU R, Brater DC, Liang WMM. PharmaIntermlttn va ContInuous Furosemide in CHF (Lahav et 8/)

15 16 17 18 19

codynamic analysis of the furosemide-probenecid interaction in man. Kidney Int 1979; 16:187-95 Odlind B, Beermann B. Renal tubular secretion and effect of furosemide. Clin Pharmacol Ther 1980; 27:784-90 Drater DC, Chennavasin ~ Seiwell R. Furosemide in patients with heart failure: shift of the dose-response curves. Clin Pharmacol Ther 1980; 28: 182-86 Drater DC. Effects of probenecid on furosemide response. Clin Pharmacol Ther 1978; 23:259-65 Vermeulen A, Chadha DR. Diuretic effect of slow release furosemide in elderly patients. Eur J Clin Pharmacoll983; 24: 449-51 Dikshit K, Vyden JK, Farrester JS, Chatterjee K, Prakash R,

20

21 22 23

Swan HJC. Renal and extrarenal hemodynamic effects of furosemide in congestive heart failure after acute myocardial infarction. N Eng) J Med 1973; 288:1087-90 Francis GS, Siegel RM, Goldsmith SR, Olivary MT, Levine TB, Cohn IN. Vasoconstrictor response to intravenous furosemide in patients with chronic congestive heart failure-activation of the neurohumoral axis. Ann Intern Med 1985; 103:1-6 Kraus PA, Lipman J, Becker PJ. Acute preload effects of furosemide. Chest 1990; 98:124-28 Larsen FF. Haemodynamic effects of high or low doses of furosemide in acute myocardial infarction. Eur Heart J 1988; 9:125-31 Lief PD. Diuretics. Am Heart J 1978; 96:824

Ninth Annual Clinical Update in Pulmonary Medicine This program will be sponsored by the Department of Pulmonary Medicine, Deborah Heart and Lung Center. It will take place at the Trump Regency Hotel, Atlantic City, November 21. For information, contact Roberta Silver, Conference Administrator, Center for Bio-MedicaI Communication, Inc, 80 West Madison Avenue, Dumont, New Jersey 07628 (201:385-8080).

CHEST I 102 I 3 I SEPTEMBER, 1992

731