Kidney International, Vol. 40 (1991), PP. 302—308
Effect of salt supplementation on amphotericin B
nephrotoxicity
ALEJANDRO LLANOS, JAVIER CIEZA, JOSE BERNARDO, JUAN ECHEVARRIA, ITALO BIAGGIONI,
RAMZI SABRA', and ROBERT A. BRANCH'
Instituto de Medicine Tropical "Alexander Von Humboldt", Universidad Peruana Cayetano Heredia, Lima, Peru, and Division of Clinical
Pharmacology, Vanderbilt University, Nashville, Tennessee, USA
Effect of salt supplementation on amphotericin B nephrotoxicity. It has
been suggested that salt loading protects against amphotericin
B-in-
duced nephrotoxicity. The influence of saline loading on the nephro-
toxic response to amphotericin 8 (50 mg/dose given i.v. over 4 hr 3
x/week for 10 weeks) was assessed in two groups of ten patients each
who were diagnosed with mucocutaneous leishmaniasis. Patients were
randomized to receive either I liter of 0.9% saline or I liter of 5%
dextrose in water, administered i.v. over one hour in a double-blinded
manner, directly prior to amphotericin B administration. Renal function
was monitored on a weekly basis two days after the last dose of
amphotericin B. Baseline characteristics were similar in both groups
except for a slightly higher serum creatinine concentration (Cr) in the
saline group (0.8
0.05 vs. 0.6 0.04 mg/dl). Baseline sodium (Na)
excretion was relatively high (262 23 mmol/day in the dextrose group
and 224 17 mmollday in the saline group). None of the patients
sustained an increase in Cr to values greater than 1.7 mg/dl. Although
mean Cr remained within normal, there was a significant difference
between the two groups over the ten week period, with the dextrose
group sustaining a significant increase in Cr and the saline group
remaining unchanged. Serum potassium (K) levels fell in both groups
necessitating oral K supplementation. The saline group required signif-
icantly greater amounts of K supplementation to maintain a normal
serum K. Amphotericin B caused a rapid reduction in the acidification
ability of the kidney in response to an ammonium chloride load. Under
these conditions, the saline group had a poorer ability to acidify the
urine. Urine volumes were not different between the two groups, and
the specific gravity of the urine sustained a significant and similar
decrease in the two groups. Neither plasma trough amphotericin B
levels nor tissue concentrations at the site of the lesion were different
between the two groups, suggesting that the effect of salt is probably not
on the basis of pharmacokinetic changes. This study supports the
hypothesis that salt loading confers protection against reductions in
renal function by amphotericin B, but does so at the expense of
enhancing K loss. The lower overall incidence of nephrotoxicity in
comparison to prior published experience could be due to racial
variation in responsiveness, the high baseline salt intake, or lack of
coadministration of other nephrotoxic drugs.
Nephrotoxicity has been recognized as a serious and com-
mon complication of amphotericin B use since the introduction
Present address is: Center for Clinical Pharmacology, University of
Pittsburgh School of Medicine, 623 Scaife Hall, Pittsburgh, Pa 15261,
USA,
Received for publication October 9, 1990
and in revised form February 25, 1991
Accepted for publication April 2, 1991
© 1991 by the International Society of Nephrology
of this drug in the 1950's. Early reports indicated that as many
as 80 to 90% of patients who received the drug developed some
degree of renal dysfunction [1—3]. The clinical presentation of
nephrotoxicity includes azotemia, renal tubular acidosis, im-
paired concentrating ability, and renal sodium, potassium and
magnesium wasting, with consequent dehydration, hypokale-
mia and hypomagnesemia. The decrease in GFR may be so
severe as to necessitate discontinuation of therapy, leading to
progression of the underlying disease and extension of the
hospital stay. Therefore, any maneuver that decreases the
frequency and/or severity of amphotericin B-associated neph-
rotoxicity will aid the therapeutic use of this drug. This is
especially true considering the increasing frequency of fungal
infections, which is now reported at between 5 and 12% of
hospitalized patients [4].
Several manipulations have been proposed to try and mini-
mize amphotencin B-induced nephrotoxicity. Mannitol co-
administration was once suggested as protective based on
anecdotal, observational reports [5, 6]. A small prospective and
randomized trial, however, did not support a protective effect
[7]. More recent reports suggest that liposomal preparations of
amphotericin B may have a wider margin of safety [8—10], but
the drug has not been approved for use yet, and more definitive
statements on its superiority to conventional formulations await
larger controlled trials.
The maneuver that has received attention involves adminis-
tration of a salt load in association with the amphotericin B
dose. Unfortunately, there has not been a prospective, random-
ized, placebo-controlled trial that would provide a convincing
argument for the routine use of such a protective measure. All
previous studies have been either case reports [11], retrospec-
tive studies [12, 13] or prospective non-controlled studies [12,
14]. Major difficulties are faced in attempting to interpret the
results of these studies, namely, that they were uncontrolled,
and that certain bias may have been introduced on the basis of
selection of patients, their underlying diseases, and the admin-
istration of other drugs. The randomized, placebo controlled
study would overcome these difficulties and provide a more
solid basis to make an informed clinical decision.
Studies of the influence of salt loading on amphotericin B
nephrotoxicity in the population of patients receiving the drug
in the U.S.A. have presented some difficulties in the evaluation
of such an intervention, since they involved patients who were
302