- Email: [email protected]
The effect of lactulose on urea metabolism and nitrogen excretion in cirrhotic patients
GASTROENTEROLOGY
LIVER
PHYSIOLOGY
AND
77:518-523,1979
DISEASE
The Effect of Lactulose on Urea Metabolism and Nitrogen Excretion in Cirrhotic Patients FREDRICK
L. WEBER, JR.
Medical Service, Veterans Administration Hospital and the Department of Medicine, University of Kentucky Medical School, Lexington, Kentucky
The mechanism of action of lactulose is not known, although in vitro evidence suggests that lactulose may increase ammonia utilization and decrease ammonia production by gut flora. lf these changes occur in patients, they should be reflected in altered urea metabolism and nitrogen excretion. In seven studies conducted in 6 cirrhotic patients, the effects of lactulose on the kinetics of urea metabolism and nitrogen excretion were determined. Lactulose caused a fall in urea production t--24%, P < 0.005) that was reflected in a decrease in both urea degradation and urinary urea excretion. Likewise, lactulose caused a decrease in the total body urea pool. The fall in urinary urea was accompanied by a large (two- to threefold) increase in stool nitrogen that was of a similar magnitude of the fall in urinary urea. Although urea degradation fell after lactulose, the intestinal (extrarenal) clearance of urea did not, indicating that the fall in urea degradation was due to the observed fall in the urea pool. The results indicate that: (a) Lactulose decreases urea production by increasing the fecal output of nitrogen, a finding compatible with altered ammonia metabolism by gut flora. (b) Lactulose decreases urea degradation, although this effect is primarily the result of a fall in
Received February 8,1979. Accepted April 191979. Address reprint requests to: Frederick L. Weber, Jr., M.D., Medical Service, Veterans Administration Hospital, Cooper Drive Division, Lexington, Kentucky 40506. A portion of this work has appeared in abstract form in Gastroenterology 75:992, 1978, and was presented at the annual meeting of the American Association for the Study of Liver Diseases, November 7, 1978, Chicago, Illinois. This work was supported by the Veterans Administration. I wish to thank Dr. M. Walser for advice; Kathy Fresard, Gary Veach, and Robert Locke for technical assistance; and the nursing and dietetic staff of the Special Diagnostics and Treatment Unit. 0 1979 by the American Gastroenterological Association 0016-5085/79/090518-06$02.00
the urea pool and cannot be attributed to on inhibition of urea breakdown in the gut lumen. The synthetic dissacharide lactulose is an effective agent in the treatment of portal-systemic encephalopathy. The beneficial action of lactulose has been related to its effectiveness in lowering blood ammonia levels,‘-4 although the mechanisms by which lactulose acts to bring out this result remain unknown. Hypotheses that have been refuted were that lactulose might alter bacterial flora in the gut, reducing the number of urea-splitting organisms,5.6 or that because of an acid intracolonic environment, large quantities of ammonium might be trapped and excreted in the stoo1.4.7 However, the recent in vitro studies of Vince et al.‘.’ have suggested other mechanisms by which lactulose might reduce gut ammonia production: (a) In incubated stool samples, lactulose appeared to alter the metabolism of gut flora, increasing ammonia incorporation and diminishing ammonia production.” (b) Bacterial urease activity was reduced under acidic conditions comparable with those seen in the feces after lactulose administration.” It has not been demonstrated whether these observations are relevant to the efficacy of lactulose in patients. The purpose of this study was to determine the effects of lactulose on gut ammonia production in patients. This was done by evaluating changes in the kinetics of urea metabolism including total production rate and degradation rate as well as total nitrogen excretion in urine and feces, Because urea synthesis is in part dependent upon ammonia produced in the gut, the total urea production rate was used as an indirect measure of gut ammonia production. A portion of the ammonia produced in the gut is in turn derived from the urea degraded by intestinal flora,l’ hence the effects of lactulose on urea breakdown are particularly relevant.
September
Table
METABOLIC
1979
1. Clinical
Laboratory
Data on the 6 Cirrhotic
Aspartate amino transferase (I.U./liter)
Patient No. 1
Alkaline phosphatase (I.U./liter)
38 20 46 24 19 40 49 (10-40)
2 3 4
5 6” 6h Normal range I’First study.
87 131 119 96 171 88
Patients
(30%0)
Albumin
(mg/dJ)
(g/d]) 3.6 4.4 3.0 4.4 5.0 3.7 (3.E.O)
and Materials
Patients Seven studies were conducted in 6 male patients with cirrhosis caused by alcohol whose age ranged from 48 to 64 yr. Cirrhosis had been demonstrated in all patients on a prior liver biopsy. Four studies (1, 3,6a, 6b) were conducted in 3 patients who had mild, chronic portal-systemic encephalopathy (Table 1). All patients denied having ingested alcohol for at least 3 mo before entering the study. They gave written consent to participate after being informed of risks and potential inconveniences. Psychiatric consultation was obtained to determine that informed consent could be given by patients with mild encephalopathy.
Protocol All subjects were admitted to the Special Diagnostics and Treatment Unit of the Veterans Administration Hospital, Lexington, Kentucky. Individual food components in each subject’s diet were from a constant source. The diets consisted of 1 g protein/kg body weight/
40
5 f U
20 -
r.0.99
:
1
1
200
1
1
600
MINUTES
Figure
(se4 15.2 12.6 16.6 12.4 12.4 13.7 12.8 (10-14)
Creatinine
Urea nitrogen
(mg/dl)
(mg/dJ)
0.9
1.2 1.5 2.2 1.5 1.4 1.4 (0.8-1.4)
16 15 20 25 20 21
Creatinine clearance (ml/min/ 1.73Mz) 111 85 60 46 73 51 61
1000
FROM
do0
INJECTION
day and adequate additional calories as fat and carbohydrate to maintain body weight. Protein and calorie intake was recorded after each meal. A i’-day control period consisted of a 3-day equilibration period on the diet followed by 4 days of urine and stool collections. Lactulose was then given for 7 days, with samples being collected during the last 6 days of the period. Urine was collected over each 24-hr period and analyzed for urea, creatinine, and total nitrogen. Stool, analyzed for total nitrogen, was collected over periods ranging from 1 to 3 days. Lactulose syrup (67% wt/vol; kindly provided by Dr. Milton Lieberthal, Merrell National Laboratories, Cincinnati, Ohio) was begun in a dose of 20 g P.O. each hour until the first loose bowel movement, and subsequently the dose was individualized (40-80 g/day in divided doses) so that patients had two to four semiformed bowel movements per day. Stool pH fell from a mean of 6.6 to 5.5 (P < 0.05) after the introduction of lactulose. A new steady state in urinary urea output was attained after 1 day of lactul?se. Liver tests (Table I), serum electrcllytes, creatinine, creatinine clearance, renal urea clea,,dnce. ;lnd body weight were monitored daily and showcc! no significant change throughout the studies. Three studlr,s (1, 3, 6a) were conducted while the patients wercb mamtained on a constant dose of aldactone which had tIeen given before admission (100-200 mg/day). One paiient (#3) had ascites which did not change during the study HIS ass~:ssed by abdominal girth and body weight. The kinetics of urea metabolism werp studled on day 5 of both control and lactulose periods using the procedure outlined by Walser and Bodenlos.‘” After the previous day’s stool and urine collections were completed at R:OO AM, approximately 3 PCi of urea, which had been sterilized by filtration and found to be pyrogen free, were injected. For 24 hr, each voiding was collected separately and the exact time noted. Blood was sampled hefort, ihe injection, 2 hr after the injection, and subsequently evrry 2-3 hr for 8 hr. Plasma and urine were subsequent!;; analyzed for C” urea.
1 The decline in plasma Cl4 urea before the administration of lactulose in a patient (No. 3) with ascites. In spite of the presence of ascites. there was a simple logarythmic decline in Cl4 urea. A sample for analysis of C“’ urea in ascitic fluid was taken at the time indicated (0).
Prothrombin time
’ Second study.
Methods
5
519
OF LACTLJLOSE
Studied
Total bilirubin
4.7 0.6 1.7 0.6 0.6 2.6 2.1 (0.2-1.0)
EFFECTS
Analytical
plasma
Methods
Cl4 urea was determined m aliquots of urine and after incubation with 75 units of jack bean urease
520
Table
WEBER
2. Effect
GASTROENTEROLOGY
of Lactulose
on Urea Metabolism
Urea nitrogen production (mg/% hr)” IJrea nitrogen appearance (mg/24 hr)” Urea nitrogen degradation (mg/24 hr)b Urea nitrogen pool (mg)h Plasma urea nitrogen (mg/liter) Urea space (ml)” Renal urea clearance (ml/24 hr)” Extra renal urea clearance (ml/24 hr)” ClSignificance level of the difference between was 89 +- 5 in these patients. ’ P z 0.05.
in Cirrhotic
Lactulose
164+12
123 + 13 9 26 24 23 84 42
and lactulose
(Type IX, Sigma Chemical Co., St. Louis, MO.), and trapping the liberated CZ40, in center well vials containing Hyamine (New England Nuclear, Boston, Mass.).” Radioactivity was determined in a Packard Tri Carb liquid scintillation counter (Packard Instrument Co., Downers Grove, Ill.) utilizing an internal standard to correct for quenching. Recovery of added C” urea was 96-101% when 1 ml of plasma was incubated with 1 ml of 0.067 M potassium phosphate buffer (pH 7.0), but recoveries from urine were poor using the same technique. Hence, 0.2 ml of urine was incubated in 1 ml of 0.5 M Tris buffer (pH 8.3) which yielded 95-101% recovery of added (Y4 urea. Plasma and urine samples were usually run separately, in which case a portion of the injected C’* urea solution was determined each time, and sample counts were corrected for any difference in the injection standards (maximum 4% variation). Plasma and urine C’” counts were no higher than background immediately preceding the second Cl4 urea in-
jection. Total nitrogen was determined in stool homogenates and urine on a Coleman Model 29 Nitrogen Analyzer (Coleman Instrumccnt Co., Maywood, Ill.). Recovery of nitrogen from urea. creatinine, and ornithine added to stool homogenates and urine ranged from 99 to 100%. A portion of each stool homogenate was lyophilized for a determination of the dry weight. Urea and creatinine in urine and serum were determined by automated methods in the Technicon SMA-6 and SMA-12 instruments, respectively.
Calculations Determination of urea metabolism must take into account the two primary fates of urea synthetized in the body: (a) urea that is catabolized in the gut; (b) urea that is excreted in the urine. The term “urea appearance” is applied to the urinary urea output when this figure is corrected for any change in the total body urea pool over the time of study.” In patients without renal failure, as in our patients. the latter correction is relatively minor. The kinetics of urea metabolism were determined by the The quantity of C“’ urea injected must be corrected for excess excretion of isotope during equilibration which in turn requires a determination of the urinary delay time.‘O These latter calculations resulted in relatively minor adjustments in the quantity of injected isotope which effectively equilibrated (4 5%). ??
(n = 7 Studies)
Control
119 + 45+ 262 + 221+ 562 f 570 f 202 f control
Patients
periods
Vol. 77, No. 3
Change from control 9
-4Of
96 * 12 26-c 9 219 + 23 187 & 17 543 f 20 555 f 70 149 f. 45
-24 -17k -43 -35 -19 -15 -52
by paired
t-test.
b Expressed
PC”
9
0.005
5 6 +- 8 + 10 + 23 + 35 f 30
0.005 0.05 0.005 0.02 N.S.” N.S.” N.S.”
+
per kilogram
body weight which
method of Walser and Bodenlos.“’ Using this method, the disappearance curve of Cl4 urea is plotted as a semi-log function vs. time (correlation coefficients in these studies were all < 0.98). In patient 3, who had ascites, there was a simple logarithmic decline in plasma Cl4 urea vs. time (Figure 1) as observed in normal patients”’ and our cirrhotic patients without ascites. The distribution volume of urea (urea space) is determined by dividing the quantity of Cl4 urea injected by the extrapolated concentration of Cl4 urea in plasma water at the time of injection.* The urea space multiplied by the concentration of urea in plasma water yields the calculated total body urea pool. The urea production rate is calculated by multiplying the urea pool times the slope of Cl4 urea disappearance in plasma water. Urea degradation is then given by the difference between total urea production and urea appearance. Extra-renal (intestinal) clearance of urea was calculated by dividing the urea degradation rate by the urea concentration in plasma water.” The nonurea space, equivalent to body solids, is the difference between body weight and urea space. If the techniques were reliable, the nonurea space should show little change in these patients over 2 wk, when ingesting appropriate calories to maintain body weight. In fact, the change in nonurea space between studies was 4.3 f 1.1% when calculated without regard to sign or +0.4 +- 2.0% if the sign was considered. Values are expressed as the mean f standard error of the mean. Significance was determined using the t-test for paired variables.** The two determinations of urea production in each patient were also compared after calculating the 95% confidence limits for individual values by the method of Halperin and Walse+’ utilizing a program kindly supplied by Dr. M. Walser (Department of Pharmacology and Experimental Therapeutics, Johns Hopkins University School of Medicine). To normalize the results, they have been expressed per kilogram body weight.
Results Effect of Lactulose Metabolism (Table
on Kinetics of Urea 2)
Patients were in a steady state in regard to determination of the total body urea pool during the control period and again 4 days after the in-
Scptcmber
METABOLIC
1979
EFFECTS
521
OF LACTULOSE
5.0 g/day;
P < 0.005). Stool nitrogen averaged 2.7 g/ day while the patients were on lactulose, and excretion of as much as 4 g/day was seen in 2 patients. Conversely, there was a fall in total urinary nitrogen excretion, which could be entirely accounted for by the fall in urinary urea. The increment in fecal nitrogen was quantitatively similar to the fall in urinary nitrogen, and because nitrogen intake remained unchanged, lactulose had no effect on nitrogen balance.
140 1
.i_!_l ** *
**
Discussion DAY OFCONTROL PERIOD
Figure
DAY OF LACTULOSE PERIOD
2. Change in the urea pool over the period of study (n = 7). Significant changes from the last day of the control period, determined by paired t-test, are indicated: ‘P < 0.005, ‘*P < 0.001.
troduction of lactulose (Figure 2); however, after lactulose, the urea pool fell. Lactulose caused a significant fall (-24%; P < 0.005) in the urea production rate. In 5 of 7 patient studies there was no overlapping of 95% confidence limits for urea production rates obtained before and after lactulose. The fall in urea production was reflected in decreased rates of both urea appearance and urea degradation, the two components of urea production. Eighty-two percent of the fall in urea appearance resulted from the change in urinary urea excretion and the remainder from decrements in the urea pool. As would be expected, the plasma urea concentration also fell; however, there was no change in urea space or the renal clearance rate of urea. Although lactulose caused a fall in the urea degradation rate, there was not a significant fall in the extrarenal clearance of urea. The latter indicates that the change in degradation rate primarily resulted from a fall in the plasma urea concentration. During control periods, patients with chronic PSE tended to have lower rates of urea production and higher rates of urea degradation and extrarenal urea clearance than patients without encephalopathy, but there was overlap between these two groups and no significant differences were found. There was also no apparent difference between the two groups in their response to lactulose as measured by the kinetics of urea metabolism. Effect of Lactuiose (Table 3)
on Nitrogen
Excretion
After lactulose, stool nitrogen excretion increased two- to threefold associated with significant increases in total stool weight (149 + 44 to 441 it- 59 g/day; P < 0.001) and stool solids (6.1 + 2.1 to 26.5 &
In these cirrhotic patients, lactulose caused a fall in urea production leading to a reduction in the total body urea pool. The fall in urea production was reflected in lower rates of both urea degradation and urea appearance. The fall in urea appearance was accompanied by a quantitatively similar rise in total stool nitrogen, which represented a two- to threefold increase over control levels. It would be very unlikely that ammonia accounted for such a large increment in stool nitrogen, because others have observed little or no change in fecal ammonia after lactulose.4~7 After lactulose, ammonia nitrogen entering the portal system fell by an average of 3.6 g/day based on the fall in total urea production. It is not possible to determine what fraction of total gut ammonia this represents, because the proportional contribution of gut ammonia to urea production is not known. However, approximately 2.5 g/day of ammonia nitrogen still entered the portal system after lactulose because of urea degradation. The quantities of ammonia arising in the gut from intestinal metabolism
Table 3. Components of Nitrogen Balance After Lactulose” (n = 7 Studies)
Before
and
P <‘J Nitrogen
intake
Stool nitrogen
Urine
Urine
nitrogen
urea nitrogen
Nitrogen
balance
Control
146k
Lactul0sc: Change’ Control Lactulosc Change’ Control Lactulose
140 f 10
Change’ Control Lactulose Change’ Control
8
--fi+
4
13f
3
30f +1a+-
4 :s
N.S."
0.001
125 *12
lOH+- 11 -172
5
118 * 14 98 + 12 -2Of 5
0.02
0.01
+5*11
Lactulose
+2+12
Change”
-3f
5
NS.”
” Values arc expressed as mg of nitrogen per 24 hr per kilogram body weight. ” Significance level of the difference between control and lactulose periods by paired t-test. ’The difference between lactulose and control periods. ” P > 0.05.
522
WEBER
GASTROENTEROLOGY
UREA APPEARANCE
Figure 3. Basic metabolic pathways
of urea in gut and liver.
of endogenous glutamine may be even larger based upon our observations in dogs.“’ From these patient studies, certain inferences can be made concerning the mechanism of action of lactulose after a consideration of basic aspects of urea metabolism in gut and liver (Figure 3). Urea production or synthesis, which is the sum of urea appearance and degradation rates, is dependent in part on the quantity of substrate nitrogen reaching the liver. Ammonia has been assumed to be an important substate in this regard, and the gut an important source of systemic ammonia. Hence, if the quantities of ammonia entering the portal system were to fall, urea production would be expected to fall. If more nitrogen were functionally trapped in the bowel lumen either as ammonia or precursors of ammonia, one would expect a rise in total stool nitrogen excretion and a fall in the urea pool and the urea appearance rate, the latter two reflecting a fall in urea production. Urea degradation might fall in this circumstance in proportion to the fall in the urea pool, but extrarenal urea clearance should not change. The studies of Gibson et a1.15 have demonstrated that when the urea pool is changed by altering dietary protein, the urea degradation rate changes, but extrarenal urea clearance is unaltered. The effects of lactulose were consistent with a functional trapping of ammonia precursors in the gut lumen. On the other hand, if bacterial urease activity alone were inhibited by lactulose, less urea would be degraded. This should have resulted in decreased rates of both extrarenal urea clearance and urea production, as has been seen in uremic” and normal” subjects after treatment with antibiotics. In the lat-
Vol. 77, No. 3
ter studies, antibiotics did not alter the urea pool or the urea appearance rate, presumably because nitrogen originating from urea degradation was largely resynthesized into urea. In contrast, lactulose had no consistent effect on extrarenal urea clearance, although the urea pool and the urea appearance rate did fall. The reduction in urea degradation could be accounted for by a fall in the urea pool. Hence, these data did not demonstrate an inhibition of urease activity with lactulose, even though the pH of these patients’ stools fell to 5.47 f 1.04, and an even lower pH was probably attained within the ascending colon.18 At a pH of 5.5, the urease activity of many bacteria is inhibited in vitro.Y Nevertheless, lactulose might have inhibited other bacterial enzyme systems that break down nitrogenous substates to form ammonia. Such a mechanism has been shown in a fecal incubation system” and would be consistent with the effects of lactulose in these patients. Although it was not a primary goal of this study, urea metabolism in these cirrhotic patients can be compared with values in the literature for normal subjects. In these cirrhotic patients, the mean urea production rate was 1.37 g/hr (1.26 in those with PSE), which was no different from mean rates re(1.19-1.69 g/hr) on similar ported in normal subjects protein intakes.‘“,‘“.‘7 Our findings cannot be compared with the data of Rudman et al.,” which demonstrated a reduction in the maximal rate of urea synthesis in cirrhotic subjects. In their study, patients ingested at least two-fold greater quantities of protein, and urea degradation had been largely abolished with neomycin. Urea degradation rates in the patients with PSE were no different (0.26-0.53 g/hr) than those been reported for normal subjects (0.29-0.46 g/hr),‘“.‘“.17 but the values were somewhat higher than in the cirrhotic group without PSE. Furthermore, in patients greater with PSE, urea degradation constituted portion of total urea production (36%) than in patients without PSE (17%). By comparison, mean figures in studies of normal patients have ranged from Studies of urea metabolism in cir21 to 27Y~‘“.“.‘~ rhotic patients have previously not quantitated urea increased urea degradegradation rates,‘Y.20 although dation rates have been reported with small bowel a condition that may be present in cirovergrowth,‘7 rhotic patients.“,” These findings have indicated that the major action of lactulose was functionally to trap nitrogen in the bowel. The large increment in stool nitrogen caused by lactulose was likely to be in the form of potential precursors of ammonia rather than ammonia itself, since previous studies have not demonstrated an increase in fecal ammonia after lactulose.4.7 In this regard, the data were consistent with
September
1979
those of Vince et al.,” who found that under in vitro conditions, lactulose and other sugars caused increased uptake and reduced appearance of ammonia in fecal homogenates. This effect was presumably due to alterations in bacterial metabolism. Alternatively, the findings of this study were consistent with a cathartic effect whereby, owing to a decreased transit time, contact between bacteria and ammonia precursors would be limited and increased excretion of nitrogen would result. However, most studies in which appropriate control observations have been made suggest that lactulose is more effective in lowering peripheral ammonia levels and treating PSE than other cathartics.Z.4.29.*4 The finding that lactulose did not alter nitrogen balance relates to patients studied on a normal protein intake. On a restricted protein intake, urinary urea would be expected to fall. Then, when placed on lactulose, these patients might be in negative nitrogen balance if large fecal nitrogen losses were to occur without a further reduction in urinary urea.
References ZR, et al: Comparison of lac1. Conn HO, Leevy CM, Vlahcevic tulose and neomycin in the treatment of chronic portal systemic encephalopathy. Gastroenterology 72:573-583,1977 SG, Floch MH. Conn HO: Lactulose in the treat2. Elkington ment of chronic portal-systemic encephalopathy. N Engl ) Med 281:408-412,1969 F, Goldstein H, Boyle JD: A controlled clinical trial 3. Simmons of lactulose in hepatic encephalopathy. Gastroenterology 59:827-832,197O JE, Fenton JCB, et al: Some observa4. Zeegan R, Drinkwater tions on the effects of treatment with lactulose on patients with chronic hepatic encephalopathy. Q J Med 39:245-263, 1970 ac5. Conn HO, Floch MH: Effects of lactulose and lactobacillus idophilus on the fecal flora. Am J Clin Nutr 23:1588-1594,197O JE, et al: The effect of lactu6. Vince A, Zeegen R, Drinkwater lose on the faecal flora of patients with hepatic encephalopathy. J Med Microbial 7:163-168,1974
METABOLIC
EFFECTS
OF LACTULOSE
L, Down PF, Murison J, et al: Faecal ammonia 7. Agostini pH during lactulose administration in man: comparison
523
and with
other cathartics. Gut 13:859-866,1972 on am8. Vince A, Killingley M, Wrong OM: Effect of lactulose monia production in a fecal incubation system. Gastroenterology 74:544-549,1978 production by 9. Vince A, Dawson AM, Park N, et al: Ammonia intestinal bacteria. Gut 14:171-177, 1973 in man. J Clin Invest 10. Walser M, Bodenlos LJ: Urea metabolism 38:1617-1626,1959 11. Walser M, Coulter AW, Dighe S, et al: The effect of keto analogues of essential amino acids in severe chronic uremia. J Clin Invest 52:678-690,1973 GW, Cochran WG: Statistical Methods. Sixth Edi12. Snedecor tion. Ames, Iowa, Iowa State University Press, 1967 M, Walser M: The reliability of estimated rates of 13. Halperin production in simple turnover experiments. Arch Biochem Biophys 70:141-149.1957 14. Weber FL, Jr., Veach GL: The importance of the small intestine in gut ammonia production in the fasting dog. Gastroenterology 77:235-240,1978 15. Gibson JA, Park NJ, Sladen GE, et al: The role of the colon in urea metabolism in man. Clin Sci Mol Med 50:51-59, 1976 16. Mitch WE, Lietman PS, Walser M: Effects of oral neomycin and kanamycin in chronic uremic patients: I. Urea metabolism. Kidney Int 11:116-122.1977 17. Jones EA. Smallwood RA. Craigie A, et al: The enterohepatic circulation of urea nitrogen. Clin Sci 37:825-836, 1969 18. Brown RL, Gibson JA, Sladen GE, et al: Effects of lactulose and other laxatives on ileal and colonic pH as measured by a radiotelemetry device. Gut 15:999-1004. 1974 D, DiFulco T, Galambos JT, et al: Maximal rates of 19. Rudman excretion and synthesis of urea in normal and cirrhotic patients. J Clin Invest 52:2241-2249,1973 after oral protein inges20. Rafoth RJ, Onstad GR: Urea synthesis tion in man. J Clin Invest 56:1170-1174,1975 SL, Levitan R: Intestinal microflora in patients 21. La1 D, Gorbach with alcoholic cirrhosis: urea splitting bacteria and neomycin resistance. Gastroenterology 62:275-279,1972 22. Martini GA, Phear EA, Ruebner B, et al: The bacterial content of the small intestine in normal and cirrhotic subjects: relation to methionine toxicity Clin Sci 16:35-51,1956 WV Jr: Encephalopathy after 23. Brown H, Trey C, McDermott portacaval shunting managed with lactulose. Am J Surg 119:132-137,197o 24. Bircher J, Haemmerli UP, Scollo-Lavizzari G: Treatment of chronic portal-systemic encephalopathy with lactulose. Am J Med 51:148-159,197l
LIVER
PHYSIOLOGY
AND
77:518-523,1979
DISEASE
The Effect of Lactulose on Urea Metabolism and Nitrogen Excretion in Cirrhotic Patients FREDRICK
L. WEBER, JR.
Medical Service, Veterans Administration Hospital and the Department of Medicine, University of Kentucky Medical School, Lexington, Kentucky
The mechanism of action of lactulose is not known, although in vitro evidence suggests that lactulose may increase ammonia utilization and decrease ammonia production by gut flora. lf these changes occur in patients, they should be reflected in altered urea metabolism and nitrogen excretion. In seven studies conducted in 6 cirrhotic patients, the effects of lactulose on the kinetics of urea metabolism and nitrogen excretion were determined. Lactulose caused a fall in urea production t--24%, P < 0.005) that was reflected in a decrease in both urea degradation and urinary urea excretion. Likewise, lactulose caused a decrease in the total body urea pool. The fall in urinary urea was accompanied by a large (two- to threefold) increase in stool nitrogen that was of a similar magnitude of the fall in urinary urea. Although urea degradation fell after lactulose, the intestinal (extrarenal) clearance of urea did not, indicating that the fall in urea degradation was due to the observed fall in the urea pool. The results indicate that: (a) Lactulose decreases urea production by increasing the fecal output of nitrogen, a finding compatible with altered ammonia metabolism by gut flora. (b) Lactulose decreases urea degradation, although this effect is primarily the result of a fall in
Received February 8,1979. Accepted April 191979. Address reprint requests to: Frederick L. Weber, Jr., M.D., Medical Service, Veterans Administration Hospital, Cooper Drive Division, Lexington, Kentucky 40506. A portion of this work has appeared in abstract form in Gastroenterology 75:992, 1978, and was presented at the annual meeting of the American Association for the Study of Liver Diseases, November 7, 1978, Chicago, Illinois. This work was supported by the Veterans Administration. I wish to thank Dr. M. Walser for advice; Kathy Fresard, Gary Veach, and Robert Locke for technical assistance; and the nursing and dietetic staff of the Special Diagnostics and Treatment Unit. 0 1979 by the American Gastroenterological Association 0016-5085/79/090518-06$02.00
the urea pool and cannot be attributed to on inhibition of urea breakdown in the gut lumen. The synthetic dissacharide lactulose is an effective agent in the treatment of portal-systemic encephalopathy. The beneficial action of lactulose has been related to its effectiveness in lowering blood ammonia levels,‘-4 although the mechanisms by which lactulose acts to bring out this result remain unknown. Hypotheses that have been refuted were that lactulose might alter bacterial flora in the gut, reducing the number of urea-splitting organisms,5.6 or that because of an acid intracolonic environment, large quantities of ammonium might be trapped and excreted in the stoo1.4.7 However, the recent in vitro studies of Vince et al.‘.’ have suggested other mechanisms by which lactulose might reduce gut ammonia production: (a) In incubated stool samples, lactulose appeared to alter the metabolism of gut flora, increasing ammonia incorporation and diminishing ammonia production.” (b) Bacterial urease activity was reduced under acidic conditions comparable with those seen in the feces after lactulose administration.” It has not been demonstrated whether these observations are relevant to the efficacy of lactulose in patients. The purpose of this study was to determine the effects of lactulose on gut ammonia production in patients. This was done by evaluating changes in the kinetics of urea metabolism including total production rate and degradation rate as well as total nitrogen excretion in urine and feces, Because urea synthesis is in part dependent upon ammonia produced in the gut, the total urea production rate was used as an indirect measure of gut ammonia production. A portion of the ammonia produced in the gut is in turn derived from the urea degraded by intestinal flora,l’ hence the effects of lactulose on urea breakdown are particularly relevant.
September
Table
METABOLIC
1979
1. Clinical
Laboratory
Data on the 6 Cirrhotic
Aspartate amino transferase (I.U./liter)
Patient No. 1
Alkaline phosphatase (I.U./liter)
38 20 46 24 19 40 49 (10-40)
2 3 4
5 6” 6h Normal range I’First study.
87 131 119 96 171 88
Patients
(30%0)
Albumin
(mg/dJ)
(g/d]) 3.6 4.4 3.0 4.4 5.0 3.7 (3.E.O)
and Materials
Patients Seven studies were conducted in 6 male patients with cirrhosis caused by alcohol whose age ranged from 48 to 64 yr. Cirrhosis had been demonstrated in all patients on a prior liver biopsy. Four studies (1, 3,6a, 6b) were conducted in 3 patients who had mild, chronic portal-systemic encephalopathy (Table 1). All patients denied having ingested alcohol for at least 3 mo before entering the study. They gave written consent to participate after being informed of risks and potential inconveniences. Psychiatric consultation was obtained to determine that informed consent could be given by patients with mild encephalopathy.
Protocol All subjects were admitted to the Special Diagnostics and Treatment Unit of the Veterans Administration Hospital, Lexington, Kentucky. Individual food components in each subject’s diet were from a constant source. The diets consisted of 1 g protein/kg body weight/
40
5 f U
20 -
r.0.99
:
1
1
200
1
1
600
MINUTES
Figure
(se4 15.2 12.6 16.6 12.4 12.4 13.7 12.8 (10-14)
Creatinine
Urea nitrogen
(mg/dl)
(mg/dJ)
0.9
1.2 1.5 2.2 1.5 1.4 1.4 (0.8-1.4)
16 15 20 25 20 21
Creatinine clearance (ml/min/ 1.73Mz) 111 85 60 46 73 51 61
1000
FROM
do0
INJECTION
day and adequate additional calories as fat and carbohydrate to maintain body weight. Protein and calorie intake was recorded after each meal. A i’-day control period consisted of a 3-day equilibration period on the diet followed by 4 days of urine and stool collections. Lactulose was then given for 7 days, with samples being collected during the last 6 days of the period. Urine was collected over each 24-hr period and analyzed for urea, creatinine, and total nitrogen. Stool, analyzed for total nitrogen, was collected over periods ranging from 1 to 3 days. Lactulose syrup (67% wt/vol; kindly provided by Dr. Milton Lieberthal, Merrell National Laboratories, Cincinnati, Ohio) was begun in a dose of 20 g P.O. each hour until the first loose bowel movement, and subsequently the dose was individualized (40-80 g/day in divided doses) so that patients had two to four semiformed bowel movements per day. Stool pH fell from a mean of 6.6 to 5.5 (P < 0.05) after the introduction of lactulose. A new steady state in urinary urea output was attained after 1 day of lactul?se. Liver tests (Table I), serum electrcllytes, creatinine, creatinine clearance, renal urea clea,,dnce. ;lnd body weight were monitored daily and showcc! no significant change throughout the studies. Three studlr,s (1, 3, 6a) were conducted while the patients wercb mamtained on a constant dose of aldactone which had tIeen given before admission (100-200 mg/day). One paiient (#3) had ascites which did not change during the study HIS ass~:ssed by abdominal girth and body weight. The kinetics of urea metabolism werp studled on day 5 of both control and lactulose periods using the procedure outlined by Walser and Bodenlos.‘” After the previous day’s stool and urine collections were completed at R:OO AM, approximately 3 PCi of urea, which had been sterilized by filtration and found to be pyrogen free, were injected. For 24 hr, each voiding was collected separately and the exact time noted. Blood was sampled hefort, ihe injection, 2 hr after the injection, and subsequently evrry 2-3 hr for 8 hr. Plasma and urine were subsequent!;; analyzed for C” urea.
1 The decline in plasma Cl4 urea before the administration of lactulose in a patient (No. 3) with ascites. In spite of the presence of ascites. there was a simple logarythmic decline in Cl4 urea. A sample for analysis of C“’ urea in ascitic fluid was taken at the time indicated (0).
Prothrombin time
’ Second study.
Methods
5
519
OF LACTLJLOSE
Studied
Total bilirubin
4.7 0.6 1.7 0.6 0.6 2.6 2.1 (0.2-1.0)
EFFECTS
Analytical
plasma
Methods
Cl4 urea was determined m aliquots of urine and after incubation with 75 units of jack bean urease
520
Table
WEBER
2. Effect
GASTROENTEROLOGY
of Lactulose
on Urea Metabolism
Urea nitrogen production (mg/% hr)” IJrea nitrogen appearance (mg/24 hr)” Urea nitrogen degradation (mg/24 hr)b Urea nitrogen pool (mg)h Plasma urea nitrogen (mg/liter) Urea space (ml)” Renal urea clearance (ml/24 hr)” Extra renal urea clearance (ml/24 hr)” ClSignificance level of the difference between was 89 +- 5 in these patients. ’ P z 0.05.
in Cirrhotic
Lactulose
164+12
123 + 13 9 26 24 23 84 42
and lactulose
(Type IX, Sigma Chemical Co., St. Louis, MO.), and trapping the liberated CZ40, in center well vials containing Hyamine (New England Nuclear, Boston, Mass.).” Radioactivity was determined in a Packard Tri Carb liquid scintillation counter (Packard Instrument Co., Downers Grove, Ill.) utilizing an internal standard to correct for quenching. Recovery of added C” urea was 96-101% when 1 ml of plasma was incubated with 1 ml of 0.067 M potassium phosphate buffer (pH 7.0), but recoveries from urine were poor using the same technique. Hence, 0.2 ml of urine was incubated in 1 ml of 0.5 M Tris buffer (pH 8.3) which yielded 95-101% recovery of added (Y4 urea. Plasma and urine samples were usually run separately, in which case a portion of the injected C’* urea solution was determined each time, and sample counts were corrected for any difference in the injection standards (maximum 4% variation). Plasma and urine C’” counts were no higher than background immediately preceding the second Cl4 urea in-
jection. Total nitrogen was determined in stool homogenates and urine on a Coleman Model 29 Nitrogen Analyzer (Coleman Instrumccnt Co., Maywood, Ill.). Recovery of nitrogen from urea. creatinine, and ornithine added to stool homogenates and urine ranged from 99 to 100%. A portion of each stool homogenate was lyophilized for a determination of the dry weight. Urea and creatinine in urine and serum were determined by automated methods in the Technicon SMA-6 and SMA-12 instruments, respectively.
Calculations Determination of urea metabolism must take into account the two primary fates of urea synthetized in the body: (a) urea that is catabolized in the gut; (b) urea that is excreted in the urine. The term “urea appearance” is applied to the urinary urea output when this figure is corrected for any change in the total body urea pool over the time of study.” In patients without renal failure, as in our patients. the latter correction is relatively minor. The kinetics of urea metabolism were determined by the The quantity of C“’ urea injected must be corrected for excess excretion of isotope during equilibration which in turn requires a determination of the urinary delay time.‘O These latter calculations resulted in relatively minor adjustments in the quantity of injected isotope which effectively equilibrated (4 5%). ??
(n = 7 Studies)
Control
119 + 45+ 262 + 221+ 562 f 570 f 202 f control
Patients
periods
Vol. 77, No. 3
Change from control 9
-4Of
96 * 12 26-c 9 219 + 23 187 & 17 543 f 20 555 f 70 149 f. 45
-24 -17k -43 -35 -19 -15 -52
by paired
t-test.
b Expressed
PC”
9
0.005
5 6 +- 8 + 10 + 23 + 35 f 30
0.005 0.05 0.005 0.02 N.S.” N.S.” N.S.”
+
per kilogram
body weight which
method of Walser and Bodenlos.“’ Using this method, the disappearance curve of Cl4 urea is plotted as a semi-log function vs. time (correlation coefficients in these studies were all < 0.98). In patient 3, who had ascites, there was a simple logarithmic decline in plasma Cl4 urea vs. time (Figure 1) as observed in normal patients”’ and our cirrhotic patients without ascites. The distribution volume of urea (urea space) is determined by dividing the quantity of Cl4 urea injected by the extrapolated concentration of Cl4 urea in plasma water at the time of injection.* The urea space multiplied by the concentration of urea in plasma water yields the calculated total body urea pool. The urea production rate is calculated by multiplying the urea pool times the slope of Cl4 urea disappearance in plasma water. Urea degradation is then given by the difference between total urea production and urea appearance. Extra-renal (intestinal) clearance of urea was calculated by dividing the urea degradation rate by the urea concentration in plasma water.” The nonurea space, equivalent to body solids, is the difference between body weight and urea space. If the techniques were reliable, the nonurea space should show little change in these patients over 2 wk, when ingesting appropriate calories to maintain body weight. In fact, the change in nonurea space between studies was 4.3 f 1.1% when calculated without regard to sign or +0.4 +- 2.0% if the sign was considered. Values are expressed as the mean f standard error of the mean. Significance was determined using the t-test for paired variables.** The two determinations of urea production in each patient were also compared after calculating the 95% confidence limits for individual values by the method of Halperin and Walse+’ utilizing a program kindly supplied by Dr. M. Walser (Department of Pharmacology and Experimental Therapeutics, Johns Hopkins University School of Medicine). To normalize the results, they have been expressed per kilogram body weight.
Results Effect of Lactulose Metabolism (Table
on Kinetics of Urea 2)
Patients were in a steady state in regard to determination of the total body urea pool during the control period and again 4 days after the in-
Scptcmber
METABOLIC
1979
EFFECTS
521
OF LACTULOSE
5.0 g/day;
P < 0.005). Stool nitrogen averaged 2.7 g/ day while the patients were on lactulose, and excretion of as much as 4 g/day was seen in 2 patients. Conversely, there was a fall in total urinary nitrogen excretion, which could be entirely accounted for by the fall in urinary urea. The increment in fecal nitrogen was quantitatively similar to the fall in urinary nitrogen, and because nitrogen intake remained unchanged, lactulose had no effect on nitrogen balance.
140 1
.i_!_l ** *
**
Discussion DAY OFCONTROL PERIOD
Figure
DAY OF LACTULOSE PERIOD
2. Change in the urea pool over the period of study (n = 7). Significant changes from the last day of the control period, determined by paired t-test, are indicated: ‘P < 0.005, ‘*P < 0.001.
troduction of lactulose (Figure 2); however, after lactulose, the urea pool fell. Lactulose caused a significant fall (-24%; P < 0.005) in the urea production rate. In 5 of 7 patient studies there was no overlapping of 95% confidence limits for urea production rates obtained before and after lactulose. The fall in urea production was reflected in decreased rates of both urea appearance and urea degradation, the two components of urea production. Eighty-two percent of the fall in urea appearance resulted from the change in urinary urea excretion and the remainder from decrements in the urea pool. As would be expected, the plasma urea concentration also fell; however, there was no change in urea space or the renal clearance rate of urea. Although lactulose caused a fall in the urea degradation rate, there was not a significant fall in the extrarenal clearance of urea. The latter indicates that the change in degradation rate primarily resulted from a fall in the plasma urea concentration. During control periods, patients with chronic PSE tended to have lower rates of urea production and higher rates of urea degradation and extrarenal urea clearance than patients without encephalopathy, but there was overlap between these two groups and no significant differences were found. There was also no apparent difference between the two groups in their response to lactulose as measured by the kinetics of urea metabolism. Effect of Lactuiose (Table 3)
on Nitrogen
Excretion
After lactulose, stool nitrogen excretion increased two- to threefold associated with significant increases in total stool weight (149 + 44 to 441 it- 59 g/day; P < 0.001) and stool solids (6.1 + 2.1 to 26.5 &
In these cirrhotic patients, lactulose caused a fall in urea production leading to a reduction in the total body urea pool. The fall in urea production was reflected in lower rates of both urea degradation and urea appearance. The fall in urea appearance was accompanied by a quantitatively similar rise in total stool nitrogen, which represented a two- to threefold increase over control levels. It would be very unlikely that ammonia accounted for such a large increment in stool nitrogen, because others have observed little or no change in fecal ammonia after lactulose.4~7 After lactulose, ammonia nitrogen entering the portal system fell by an average of 3.6 g/day based on the fall in total urea production. It is not possible to determine what fraction of total gut ammonia this represents, because the proportional contribution of gut ammonia to urea production is not known. However, approximately 2.5 g/day of ammonia nitrogen still entered the portal system after lactulose because of urea degradation. The quantities of ammonia arising in the gut from intestinal metabolism
Table 3. Components of Nitrogen Balance After Lactulose” (n = 7 Studies)
Before
and
P <‘J Nitrogen
intake
Stool nitrogen
Urine
Urine
nitrogen
urea nitrogen
Nitrogen
balance
Control
146k
Lactul0sc: Change’ Control Lactulosc Change’ Control Lactulose
140 f 10
Change’ Control Lactulose Change’ Control
8
--fi+
4
13f
3
30f +1a+-
4 :s
N.S."
0.001
125 *12
lOH+- 11 -172
5
118 * 14 98 + 12 -2Of 5
0.02
0.01
+5*11
Lactulose
+2+12
Change”
-3f
5
NS.”
” Values arc expressed as mg of nitrogen per 24 hr per kilogram body weight. ” Significance level of the difference between control and lactulose periods by paired t-test. ’The difference between lactulose and control periods. ” P > 0.05.
522
WEBER
GASTROENTEROLOGY
UREA APPEARANCE
Figure 3. Basic metabolic pathways
of urea in gut and liver.
of endogenous glutamine may be even larger based upon our observations in dogs.“’ From these patient studies, certain inferences can be made concerning the mechanism of action of lactulose after a consideration of basic aspects of urea metabolism in gut and liver (Figure 3). Urea production or synthesis, which is the sum of urea appearance and degradation rates, is dependent in part on the quantity of substrate nitrogen reaching the liver. Ammonia has been assumed to be an important substate in this regard, and the gut an important source of systemic ammonia. Hence, if the quantities of ammonia entering the portal system were to fall, urea production would be expected to fall. If more nitrogen were functionally trapped in the bowel lumen either as ammonia or precursors of ammonia, one would expect a rise in total stool nitrogen excretion and a fall in the urea pool and the urea appearance rate, the latter two reflecting a fall in urea production. Urea degradation might fall in this circumstance in proportion to the fall in the urea pool, but extrarenal urea clearance should not change. The studies of Gibson et a1.15 have demonstrated that when the urea pool is changed by altering dietary protein, the urea degradation rate changes, but extrarenal urea clearance is unaltered. The effects of lactulose were consistent with a functional trapping of ammonia precursors in the gut lumen. On the other hand, if bacterial urease activity alone were inhibited by lactulose, less urea would be degraded. This should have resulted in decreased rates of both extrarenal urea clearance and urea production, as has been seen in uremic” and normal” subjects after treatment with antibiotics. In the lat-
Vol. 77, No. 3
ter studies, antibiotics did not alter the urea pool or the urea appearance rate, presumably because nitrogen originating from urea degradation was largely resynthesized into urea. In contrast, lactulose had no consistent effect on extrarenal urea clearance, although the urea pool and the urea appearance rate did fall. The reduction in urea degradation could be accounted for by a fall in the urea pool. Hence, these data did not demonstrate an inhibition of urease activity with lactulose, even though the pH of these patients’ stools fell to 5.47 f 1.04, and an even lower pH was probably attained within the ascending colon.18 At a pH of 5.5, the urease activity of many bacteria is inhibited in vitro.Y Nevertheless, lactulose might have inhibited other bacterial enzyme systems that break down nitrogenous substates to form ammonia. Such a mechanism has been shown in a fecal incubation system” and would be consistent with the effects of lactulose in these patients. Although it was not a primary goal of this study, urea metabolism in these cirrhotic patients can be compared with values in the literature for normal subjects. In these cirrhotic patients, the mean urea production rate was 1.37 g/hr (1.26 in those with PSE), which was no different from mean rates re(1.19-1.69 g/hr) on similar ported in normal subjects protein intakes.‘“,‘“.‘7 Our findings cannot be compared with the data of Rudman et al.,” which demonstrated a reduction in the maximal rate of urea synthesis in cirrhotic subjects. In their study, patients ingested at least two-fold greater quantities of protein, and urea degradation had been largely abolished with neomycin. Urea degradation rates in the patients with PSE were no different (0.26-0.53 g/hr) than those been reported for normal subjects (0.29-0.46 g/hr),‘“.‘“.17 but the values were somewhat higher than in the cirrhotic group without PSE. Furthermore, in patients greater with PSE, urea degradation constituted portion of total urea production (36%) than in patients without PSE (17%). By comparison, mean figures in studies of normal patients have ranged from Studies of urea metabolism in cir21 to 27Y~‘“.“.‘~ rhotic patients have previously not quantitated urea increased urea degradegradation rates,‘Y.20 although dation rates have been reported with small bowel a condition that may be present in cirovergrowth,‘7 rhotic patients.“,” These findings have indicated that the major action of lactulose was functionally to trap nitrogen in the bowel. The large increment in stool nitrogen caused by lactulose was likely to be in the form of potential precursors of ammonia rather than ammonia itself, since previous studies have not demonstrated an increase in fecal ammonia after lactulose.4.7 In this regard, the data were consistent with
September
1979
those of Vince et al.,” who found that under in vitro conditions, lactulose and other sugars caused increased uptake and reduced appearance of ammonia in fecal homogenates. This effect was presumably due to alterations in bacterial metabolism. Alternatively, the findings of this study were consistent with a cathartic effect whereby, owing to a decreased transit time, contact between bacteria and ammonia precursors would be limited and increased excretion of nitrogen would result. However, most studies in which appropriate control observations have been made suggest that lactulose is more effective in lowering peripheral ammonia levels and treating PSE than other cathartics.Z.4.29.*4 The finding that lactulose did not alter nitrogen balance relates to patients studied on a normal protein intake. On a restricted protein intake, urinary urea would be expected to fall. Then, when placed on lactulose, these patients might be in negative nitrogen balance if large fecal nitrogen losses were to occur without a further reduction in urinary urea.
References ZR, et al: Comparison of lac1. Conn HO, Leevy CM, Vlahcevic tulose and neomycin in the treatment of chronic portal systemic encephalopathy. Gastroenterology 72:573-583,1977 SG, Floch MH. Conn HO: Lactulose in the treat2. Elkington ment of chronic portal-systemic encephalopathy. N Engl ) Med 281:408-412,1969 F, Goldstein H, Boyle JD: A controlled clinical trial 3. Simmons of lactulose in hepatic encephalopathy. Gastroenterology 59:827-832,197O JE, Fenton JCB, et al: Some observa4. Zeegan R, Drinkwater tions on the effects of treatment with lactulose on patients with chronic hepatic encephalopathy. Q J Med 39:245-263, 1970 ac5. Conn HO, Floch MH: Effects of lactulose and lactobacillus idophilus on the fecal flora. Am J Clin Nutr 23:1588-1594,197O JE, et al: The effect of lactu6. Vince A, Zeegen R, Drinkwater lose on the faecal flora of patients with hepatic encephalopathy. J Med Microbial 7:163-168,1974
METABOLIC
EFFECTS
OF LACTULOSE
L, Down PF, Murison J, et al: Faecal ammonia 7. Agostini pH during lactulose administration in man: comparison
523
and with
other cathartics. Gut 13:859-866,1972 on am8. Vince A, Killingley M, Wrong OM: Effect of lactulose monia production in a fecal incubation system. Gastroenterology 74:544-549,1978 production by 9. Vince A, Dawson AM, Park N, et al: Ammonia intestinal bacteria. Gut 14:171-177, 1973 in man. J Clin Invest 10. Walser M, Bodenlos LJ: Urea metabolism 38:1617-1626,1959 11. Walser M, Coulter AW, Dighe S, et al: The effect of keto analogues of essential amino acids in severe chronic uremia. J Clin Invest 52:678-690,1973 GW, Cochran WG: Statistical Methods. Sixth Edi12. Snedecor tion. Ames, Iowa, Iowa State University Press, 1967 M, Walser M: The reliability of estimated rates of 13. Halperin production in simple turnover experiments. Arch Biochem Biophys 70:141-149.1957 14. Weber FL, Jr., Veach GL: The importance of the small intestine in gut ammonia production in the fasting dog. Gastroenterology 77:235-240,1978 15. Gibson JA, Park NJ, Sladen GE, et al: The role of the colon in urea metabolism in man. Clin Sci Mol Med 50:51-59, 1976 16. Mitch WE, Lietman PS, Walser M: Effects of oral neomycin and kanamycin in chronic uremic patients: I. Urea metabolism. Kidney Int 11:116-122.1977 17. Jones EA. Smallwood RA. Craigie A, et al: The enterohepatic circulation of urea nitrogen. Clin Sci 37:825-836, 1969 18. Brown RL, Gibson JA, Sladen GE, et al: Effects of lactulose and other laxatives on ileal and colonic pH as measured by a radiotelemetry device. Gut 15:999-1004. 1974 D, DiFulco T, Galambos JT, et al: Maximal rates of 19. Rudman excretion and synthesis of urea in normal and cirrhotic patients. J Clin Invest 52:2241-2249,1973 after oral protein inges20. Rafoth RJ, Onstad GR: Urea synthesis tion in man. J Clin Invest 56:1170-1174,1975 SL, Levitan R: Intestinal microflora in patients 21. La1 D, Gorbach with alcoholic cirrhosis: urea splitting bacteria and neomycin resistance. Gastroenterology 62:275-279,1972 22. Martini GA, Phear EA, Ruebner B, et al: The bacterial content of the small intestine in normal and cirrhotic subjects: relation to methionine toxicity Clin Sci 16:35-51,1956 WV Jr: Encephalopathy after 23. Brown H, Trey C, McDermott portacaval shunting managed with lactulose. Am J Surg 119:132-137,197o 24. Bircher J, Haemmerli UP, Scollo-Lavizzari G: Treatment of chronic portal-systemic encephalopathy with lactulose. Am J Med 51:148-159,197l