GASTROENTEROLOGY 1982;82:213-7
Effects of Lactulose and Neomycin on
Urea Metabolism in Cirrhotic Subjects
FREDRICK L. WEBER, JR., KATHLEEN M. FRESARD, and
BARBARA R. LALLY
Veterans ‘Administration Medical Center and Department
Kentucky Medical School, Lexington, Kentucky
In our previous studies ofcirrhotic subjects Iactulose
caused a 25% decrease in the urea production rate
associated with a decrease in urinary urea excretion
and an increase in stool nitrogen. The decrease in
the rate of urea product@ was an indirect measure
of reduction in gut ammonia production. The pre-
sent study was designed to determine if the poorly
absorbed antibiotic neomycin had an additive effect
in reducing ammonia production when adminis-
tered in combination with Iactulose. Six stable cir-
rhotic subjects received isonitrogenous diets during
separate Iactulose and Iactulose + neomycin treat-
ment periods. The addition of neomycin to a Iaclu-
lose regimen caused a 17% reduction in the urea
production rate that was quantitatively accounted
for by a 70% reduction in the urea degradation rate.
The intestinal urea clearance rate demonstrated a poorly
paraIIe1 reduction, indicating an inhibition of bacte-
rial ureolysis. There was no evidence that neomycin
altered the efiects of Iactulose since urinary urea
excretion did not rise, fecal nitrogen remained high,
and stools remained acidic. These results demon-
strate that neomycin inhibited bacterial ureolysis
when administered with Iactulose while Iactulose
itself was metabolized and its individual effect on
nitrogen metabolism persisted. Lactulose and neo-
mycin, when administered together, had an additive
Received July 27, 1981. Accepted October 2, 1981.
Address requests for reprints to: Fredrick L. Weber, Jr., M.D.,
Department of Medicine, Case Western Reserve Medical School,
Cleveland, Ohio 44106.
This study was supported by the Medical Research Service of
the Veterans Administration. We thank Robert Locke for technical
assistance, Joyce Miller for secretarial assistance, and the Dietetic
and Nursing staff of the Special Diagnostics and Treatment Unit.
A portion of this work has appeared in abstract form (Gastroen-
terology 1980; 78:1288) and was presented at the annual meeting
of the American Association for the Study of Liver Diseases,
Chicago, Illinois on November 8, 1980.
0 1982 by the American Gastroenterological Association
0016-5085/82/020213-05$02.50
of Medicine, University of
effect in reducing gut ammonia production in cir-
rhotic subjects.
Both lactulose and neomycin are therapeutic agents
with proven efficacy in the treatment of portal-
systemic encephalopathy. With each drug clinical
benefit had been linked to a reduction in arterial
blood ammonia (l-3). However, it has been uncer-
tain whether these drugs may be used together to
produce additive effects in reducing the blood am-
monia concentration. In most individuals the syn-
thetic disaccharide lactulose is metabolized by intes-
tinal flora, a process that can be monitored by the
production of acid stools. Clinical efficacy and re-
duction in blood ammonia have been linked to
bacterial metabolism of lactulose (2). In contrast, the
absorbed antibiotic neomycin has been
thought to act by reducing ureolytic and putrefactive
activity of the enteric microflora (4). Since neomycin
kills bacteria while lactulose depends on intact bac-
terial metabolism for its beneficial effects, it might be
considered inappropriate to use these drugs in com-
bination. However, since the two drugs may interact
with different bacterial populations, they could have
additive effects in reducing intestinal ammonia pro-
duction.
Previous studies have analyzed changes in the
urea production rate to quantify changes in intesti-
nal ammonia production. Lactulose has been shown
to cause a mean reduction of 25% in the urea
production rate associated with a reduction in uri-
nary urea excretion and an increase in fecal nitrogen
(5,6). On the other hand, antibiotics including neo-
mycin have been shown to reduce the urea produc-
tion rate by inhibiting urea degradation and hence,
preventing the recycling of urea and ammonia nitro-
gen (7-9). In this study we have assessed the effects
of combination therapy with lactulose and neomycin
on gut ammonia production by measuring changes
214 WBER ET AL. GASTROENTEROLOGY Vol.82,No. 2

in urea production and degradation rates as well as Bodenlos (7). After an overnight fast, patients were inject-
intestinal urea clearance. ed intravenously with approximately 5 &i of sterile
pyrogen-free [14C]urea at 8:30A.M. They then continued to
ingest their protocol diet. Urine’and blood collections for
Methods determination of [‘*C]urea were carried out as previously
described (5).
Patients [r4C]urea was determined in plasma and urine as total
[14C] on a Packard Tri-Carb scintillation counter (Packard
Six stable patients with biopsy-proven cirrhosis Instrument Co., Inc., Downers Grove, Ill.) using an external
caused by alcohol were studied. Their ages ranged from 52 standard to correct for quenching. In urine and plasma
to 65 yr and their weights from 69 to 67 kg (mean, 78 kg). samples taken from this type of study we and others have
None of these patients had ascites, edema, encephalop- found no difference between the counts determined in
athy, or gastrointestinal bleeding. They took no drugs total [14C] and counts determined in [14C] released by
during the study other than neomycin and lactulose. All urease treatment (6,lO). Total nitrogen was determined in
patients denied alcohol ingestion for at least 6 mo before neomycin tablets, selected diets, and stool and urine
study; this was confirmed by family members. Conven- samples as previously described (5). Stool dry weight was
tional liver tests were either normal or mildly abnormal determined on lyophilized specimens. Other chemical
and remained stable throughout the study period. Initial determinations were carried out on Technicon SMA-6 and
values included serum glutamic oxaloacetic acid transami- SMA-12 instruments (Technicon Instruments Corp., Tar-
nase, 36 + 6 III/L (mean + 1 SP; normal, 10-40);alkaline rytown, N.Y.).
phosphatase, 98 f 38 IV/L (normal, 30-110);total biliru-
bin, 1.5 ? 1.3mg/dl (normal, 0.2-1.0); albumin 4.0+-0.5g/
dl (normal, 3.5-5.0). Only 1 patient had a prothrombin Data Analysis
time 2 s longer than control. Serum creatinine and creati-
The principles and calculations employed in deter-
nine clearance were 1.1a 0.2mg/dl and 84 2 13 ml/mm/
mining urea production and degradation rates have been
1.73 M2, respectively, and did’not change throughout the
previously described (5,7). The urea synthetic rate was
study period. Serum electrolytes were also normal and did
determined by analysis of the logarithmic decline in
not change.
plasma [r4C]urea specific activity after injection of a
All subjects gave written informed consent to participate
known amount of (‘4C]urea. Urea synthesized in the body
in the study. The protocol was approved by the Human
has two primary fates: excretion in the urine or degrada-
Investigations and Studies Committee of the University of
tion in the gut. Urinary urea was measured directly. This
Kentucky on January 22, 1979.
value has been termed. the urea appearance rate when
changes in the total body urea pool were added (11).
Protocol However, in these patients there were no changes in the
urea pool during individual study periods; hence, urea
Patients were studied under metabolic ward condi- appearance was equal to urinary urea excretion. The urea
tions. Diets were composed of 1 g protein/kg body wt and degradation rate was determined as the difference between
sufficient calories to maintain body weight. Weighed food the urea production rate and the urea appearance rate. The
items were taken from constant sources for each patient. extrarenal or intestinal clearance rate of urea was deter-
Patients ate under observation and generally consumed mined by dividing the urea degradation rate by the plasma
their entire diet. Any uneaten food was reweighed. Two urea concentration.
patients were studied during control, lactulose, and neo- Data in the text and tables have been presented as the
mycin-lactulose periods while the other 4 patients were mean +- standard error of the mean, Statistical significance
studied during lactulose and neomycin-lactulose periods was determined by paired t-tests (12).
only. During control periods patients received nq drugs.
Lactulose syrup (67% iyflvol) was initially given in an oral
dose of 20 g/huntil the first loose bowel movement and Results
then in individualized doses (36-80 g/day) to produce 2-4
bowel movements per day. Neomycin in a dose of 4 g/day
Urea Synthesis and Degradation
was subsequently added to the lactulose regimen. The
order of treatment periods was not randomized because it In the 2 patients who were studied during a
was uncertain how long it might have taken for the colonic control period, lactulose caused a fall in the urea
flora to have become reestablished if neomycin had been production rate, urea appearance rate, and urea pool
given during the first period. Individual study periods as has been consistently seen in previous studies
lasted from 8 to 10 days. The first 3 days were necessary
(5,6). When neomycin was added to lactulose in the
for equilibration on a drug regimen and the remaining days
current study there was a further reduction in the
for collection of stool (1-3 day collections), urine (24-h
collections), and blood (daily). After 3 days patients urea production rate that was entirely accounted for
reached a steady state in regard to blood urea concentra- by 70% mean reduction in urea degradation (Table
tion and urinary urea excretion. Urea production and 1). Individual patients demonstrated a strong corre-
degradation rates were determined on day 5,6, or 7 of each lation between reductions in urea production and
study period according to the method of Walser and urea degradation (Figure 1). Neither the urea appear-
February 1982 LACTULOSE AND NEOMYCIN 215

Table 1. Effects on Urea Metabolism after Neomycin 8.0 -

Was Added to L..actuIose

Lactulose +
Lactulose neomycin Change
70 -
Urea production (mg N)O 119 2 10 102 f 8 -17 2 4b
Urea appearance (mg rJ1” 92 2 8 94 2 8 +2&l
Urea degradation (mg rJ1” 27 * 4 9*2 -18 k 4b
Plasma urea (mg N/d11 14 5 2 14 rt3 - I
0
Urea pool (mg N/kg) 76 5 11 74 r 13 -2 -c3
d 6.0 -
Renal urea clearance
E
(ml)’ 665 + 65 695 2 74 +30 rt28
Intestinal urea clearance
(mfla 182 f 32 58 2 15 -123 t 24'
50 -
a Results are expressed per 24 hours per kilogram. b p < 0.01.
c p < 0.005. N = nitrogen.

ante rate nor the urea pool changed, indicating that I I I

the effects of lactulose were not reversed, and that
urea metabolism was not changing solely as a func-
tion of time in these subjects. The intestinal clear-
ance of urea fell by 68% in parallel with the fall in
the urea degradation rate (Table 1). The fall in
intestinal urea clearance was explained by a reduc-
tion in the urea degradation rate and the absence of a
change in plasma urea concentration.
Stool Analysis
In the 2 patients studied during a control
period, stool pH showed a characteristic fall, associ-
ated with lactulose administration, when they were
begun on the drug (Figure 2). In the other 4 patients
who were initially begun on lactulose, fecal pH was
also low. The administration of neomycin along with
lactulose had no effect on fecal pH suggesting that
there was ongoing bacterial metabolism of lactulose.
In spite of giving individualized oral doses of
4.01
CONTROL LACTULOSE NEOYKlN*
LACTULOSf
Figure 2. Mean fecal pH in individual patients during control
period, lactulose period, and lactuloseneomycin peri-
od.
lactulose to produce 2-4 bowel movements per day,
total stool weight was higher during the administra-
tion of both lactulose and neomycin (604 ?C66 g/day)
than when lactulose was given alone (402 ‘-+37 g/
day). In 3 patients it was necessary to lower the dose
of lactulose after beginning neomycin therapy. In all
6 patients the dose of lactulose was 63 + 5 g/day
when given alone and 55 + 5 g/day when given with
neomycin. Fecal solids, which averaged 55 + 4 g/day
during lactulose administration, also increased sig-
nificantly after the addition of neomycin to 74 * 8 g/
day (p < 0.05). However, when the weight of neomy-
cin (4 g/day) was subtracted from fecal solid weight,
there was no longer a significant increase in fecal
solids. Similarly, the increase in fecal nitrogen ex-
cretion caused by neomycin was not significant
when the amount of nitrogen contained in neomycin
(0.38 g/day) was subtracted (Table 2).

Total Nitrogen Intake and Nitrogen Balance

There was enough variability in dietary nitro-
rz0.978 gen intake that the small amount of nitrogen con-
m =0.990
tained in neomycin caused no significant increase in
total nitrogen intake (Table 2). Urinary nitrogen

_,oK~ , , , PC0001 excretion was unchanged when neomycin

added to lactulose, and there was no change in net
was

nitrogen balance in spite of a small increase in fecal

nitrogen excretion.
-20 -30 -40
CHANGE IN PRODUCTIONRATE
UREA NITROGEN, mg, kf’,24 hr-' Discussion

Figure . Correlation between the change in urea production rate

In previously studied patients and 2 patients
and the change in urea degradation rate after neomycin from the current study lactulose caused a consistent
was added to lactulose therapy. fall in the urea production rate, urea appearance rate,
216 WEBER ET AL. GASTROENTEROLOGY Vol. 82, No. 2

Table 2. Effects on Nitrogen Metabolism of Lactulose lopathy, and in contrast to lactulose, the saline
and Lactulose + Neomycin’ cathartic magnesium sulfate has no effect on urea
Lactulose + metabolism except for a slight reduction in the urea
Lactulose neomycin Change pool (6). There was, however, no indication that
Nitrogen intake 158 f 2 161 2 1” +4 f 3 neomycin inhibited bacterial metabolism of lactu-
Urinary excretion 109 f 10 108 k 8 -1 k 4 lose in our 6 patients. Fecal pH remained low after
Fecal excretion 33 -1- 3 38 f 20 +5 + 1.5b the introduction of neomycin indicating that there
Balance +15 f 9 +16 i: 7 +1 k 5
was ongoing production of organic acids from the
a Includes nitrogen contained in neomycin. Results are expressed catabolism of lactulose. In addition to stool acidifica-
in milligrams of nitrogen per 24 hours per kilogram. tion, a repeatedly observed effect of lactulose has
b p < 0.05. been a reduction in urinary urea excretion associated
with an increase in fecal nitrogen excretion (5,6). We
have considered these changes to be consistent with
and urea pool (Figure 3). However, lactulose alone a reduced bacterial production of urea precursors
has not had any consistent effect on intestinal urea such as ammonia from nitrogen-containing com-
clearance. The results obtained from the 6 cirrhotic pounds in the colonic lumen. In the patients de-
patients included in this study demonstrated that scribed in this study, a rise in urinary urea excretion
when neomycin was added to lactulose therapy or a fall in fecal nitrogen excretion might have
there was a fall in gut ammonia production as indicated that neomycin had reversed the effects of
assessed by a reduction in the urea production rate. lactulose. However, no such changes occurred after
This reduction in urea production was entirely ac- the addition of neomycin.
counted for by an inhibition of bacterial ureolysis as
determined by reductions in both the urea degrada-
tion rate and the intestinal clearance of urea. Fur-
thermore, there was no evidence that the effects of UREA UREA
lactulose in reducing gut ammonia production were PROWCTION APPEARANCE

reversed by the administration of neomycin. p<o.o01 p'0.001

The changes in urea metabolism caused by neomy-
cin were consistent with an interruption of the
enterohepatic recycling of urea and ammonia.
Hence, there was a reduction in the quantity of urea
degraded to ammonia and, in turn, the quantity of
urea resynthesized from ammonia. By inhibiting
bacterial ureolysis, antibiotic therapy usually causes
a reduction in the urea degradation rate and intesti-
nal clearance of urea in both normal and uremic
subjects. However, the urea pool and urinary urea
excretion have not been found to change because the INTESTINAL
amount of circulating urea has not been altered (8,9). UREA UREA
POOL CLEARANCE
This relationship might not expect to hold if antibi-
p’o 001 NS
otics significantly reversed deamination of dietary or
endogenous amino acids, and, in patients with small
intestinal bacterial overgrowth, broad spectrum anti-
biotics probably do reduce urinary urea excretion
($13). In our study, the fall in the urea production
caused by neomycin could be entirely accounted for
by an inhibition of bacterial ureolysis. The increased
catharsis caused by neomycin might have contribut-
ed to a reduction in ureolysis, but in a previous
study, catharsis alone with magnesium sulfate had i
no effect on urea degradation (6).
The clinical effectiveness of lactulose and its ef- Figure 3. Effects of lactulose on urea production, urea appear-
fects on urea and nitrogen metabolism appear to ance, urea pool, and intestinal urea clearance in pa-
tients previously studied (01 (n = 12) (5,6) and in 2
depend on its capacity to be metabolized by the patients from the current study (0). p-Values indicate
intestinal flora. Other cathartics have generally not significant changes by paired t-test. NS = no significant
been effective in treating portal-systemic encepha- change, C = control period, L = lactulose period.
February 1982
The results of this study no doubt relate to the
heterogeneity of the intestinal microflora. Certain
intestinal bacteria such as Bacteroides species can
metabolize lactulose (14) but are generally resistant
to neomycin. Other ureolytic bacteria are often sensi-
tive to neomycin (15). Hence the additive effects of
these two agents in reducing gut ammonia produc-
tion probably occurred because the effect of each
drug was mediated by different bacterial popula-
tions.
Several other observations have suggested that
these two drugs might continue to act independently
when used in combination. Conn found that neomy-
cin caused fecal pH to become neutral in 3 of 11
patients who had acidic stools after lactulose (16).
The other 8 patients had either a slight rise or fall in
fecal pH indicating that neomycin did permit ongo-
ing bacterial synthesis of organic acids from lactu-
lose. Gilat et al. (17) found that neomycin caused no
inhibition of lactulose metabolism by bacteria as
assessed by breath hydrogen production. Further-
more, Murphy and Calloway (18) found that
neomycin had no consistent effect on intestinal gas
production after bean ingestion, demonstrating that
neomycin permitted ongoing bacterial metabolism of
another carbohydrate. Finally, Pirotte et al. (19) 10. Walser M. Urea metabolism in chronic renal failure. J Clin
reported that a combination of neomycin and lactu-
lose was more effective than either agent alone in
reducing the blood ammonia concentration in 9
cirrhotic patients.
Because of the small number of patients included
in our study, we cannot provide any assurance that
experience with a larger population will not disclose
a definite group of patients who fail to benefit from
the combined use of lactulose and neomycin. Data
obtained by Conn suggest that 27% of patients would
not have a favorable clinical effect if the drugs were
used in combination (16). Presumably such patients
would lack lactulose metabolizing bacteria that were
resistant to neomycin. If neomycin were to be given
to a patient receiving lactulose, it would be impor-
tant to monitor stool pH and/or blood ammonia in
addition to the patient’s clinical status to determine
whether one was justified in continuing both drugs.
Further data may be needed to evaluate the clini-
cal efficacy of a combined regimen of lactulose and
neomycin in the treatment of portal-systemic en-
cephlopathy. However, this study establishes a ratio-
nale for the combined use of these drugs.
LACTULOSE AND NEOMYCIN 217
References
1. Bircher J, Haemmerli UP, Scollo-Lavizzari G. Treatment of
chronic portal-systemic encephalopathy with lactulose. Am J
Med 1971;51:148-59.
2. Corm HO, Leevy CM, Vlahcevic ZR, et al. Comparison of
lactulose and neomycin in the treatment of chronic portal-
systemic encephalopathy: a double-blind controlled trial.
Gastroenterology 1977;72:573-83.
3. Atterbury CE, Maddrey WC, Conn HO. Neomycin-sorbitol
and lactulose in the treatment of acute portal-systemic
encephalopathy. A controlled, double-blind clinical trial. Am
J Dig Dis 1978;23:398-406.
4. Fisher CJ, Faloon WW. Blood ammonia levels in hepatic
cirrhosis. Their control by the oral administration of neomy-
cin. N Engl J Med 1957;256:1030-5.
5. Weber FL, Jr. The effect of lactulose on urea metabolism and
nitrogen excretion in cirrhotic patients. Gastroenterology
1979;77:518-23.
6. Weber FL, Jr, Fresard KM. Comparative effects of lactulose
and magnesium sulfate on urea metabolism and nitrogen
excretion in cirrhotic subjects. Gastroenterology 1981;
80:994-a.
7. Walser M, Bodenlos LJ. Urea metabolism in man. J Clin Invest
1959;38:1617-26.
8. Mitch WE, Lietman PS, Walser M. Effects of oral neomycin
and kanamycin in chronic uremic patients: I. Urea metabo-
lism. Kidney International 1977;11:116-22.
9. Jones EA, Smallwood RA, Craigie A, Rosenoer VM. The
enterohepatic circulation of urea nitrogen. Clin Sci
1969;37:825-38.
Invest 1974;53:1385-92.
11. Walser M, Coulter AW, Dighe S, Crantz F. The effect of keto
analogues of essential amino acids in severe chronic uremia. J
Clin Invest 1973;52:878-90.
12. Snedecor GW, Co&ran WG. Statistical methods. 8th ed.
Ames, Iowa: Iowa State University Press, 1987.
13. Jones EA, Craigie A, Tavill AS, et al. Protein metabolism in
the intestinal stagnant loop syndrome. Gut 1968;9:466-9.
14. Vince AJ, Burridge SM. Ammonia production by intestinal
bacteria: the effects of lactose, lactulose and glucose. J Med
Microbial 1980;13:177-91.
15. La1 D, Gorbach SL, Levitan R. Intestinal microflora in patients
with alcoholic cirrhosis: urea splitting bacteria and neomycin
resistance. Gastroenterology 1972;62:275-9.
16. Conn HO, Lieberthal MM. Lactulose and neomycin: com-
bined therapy. In: The hepatic coma syndromes and lactu-
lose. Baltimore: Williams and Wilkins Co., 1979:349-5.
17. Gilat T, Ben Hur H, Gelman-Malachi E, et al. Alterations of
the colonic flora and their effect on the hydrogen breath test.
Gut 1978;19:602-5.
18. Murphy EL, Calloway DH. The effect of antibiotic drugs on
the volume and composition of intestinal gas from beans. Am
J Dig Dis 1972;17:639-42.
19. Pirotte J, Guffens JM, Devos J. Comparative study of basal
arterial ammonemia and of orally-induced hyperammonemia
in chronic portal systemic encephalopathy treated with neo-
mycin, lactulose and an association of neomycin and lactu-
lose. Digestion 1974;10:435-44.