- Email: [email protected]
Lactose enemas plus placebo tablets vs. neomycin tablets plus starch enemas in acute portal systemic encephalopathy
GASTROENTEROLOGY
LIVER
PHYSIOLOGY
AND
1981;81:101-6
DISEASE
Lactose Enemas Plus Placebo Tablets vs. Neomycin Tablets Plus Starch Enemas in Acute Portal Systemic Encephalopathy A Double-Blind Randomized
Controlled Study
MISAEL URIBE, JESIjS MORENO BkRTHIER, HILEL LEWIS, JUAN M. MATA, JOSE G. SIERRA, GUILLERMO GARCfA-RAMOS, JAVIER RAMfREZ ACOSTA, and MARGARITA DEHESA, with the Technical Assistance of ELBA GALVAN and SARA ALVAREZ The Liver Unit, Instituto National de la Nutricibn, and the Gastrointestinal Hospital General IMSS MBxico City, MBxico
A randomized, double-blind comparison of lactose enemas plus placebo tablets vs. starch enemas plus neomycin tablets was performed on 18 patients with acute portal systemic encephalopathy. Ten patients received starch enemas (10%;2000 ml t.i.d.) plus neomycin tablets and 8 patients received lactose enemas (20%;1000 ml Cd.) plus placebo tablets. A significant mental state improvement was demonstrated in the group of patients treated with starch enemas-neomycin tablets (p c 0.05) and in the group of patients treated with lactose enemas-placebo tabBoth treatments significantly imlets (p < 0.025). proved the frequency of asterixis, ammonia blood levels, and electroencephalograms. In addition, patients treated with lactose enemas showed significant improvement in number-connection test times (p < O.O2), and their stools showed a mQre acid pH (p c 0.05). No side effects were evident with either treatment. Lactose enemas are a safe and effective treatment for acute portal systemic encephalopathy. Received September 3, 1980. Accepted February 7, 1981. Address requests for reprints to: Dr. Misael Uribe, Instituto Nacional de la Nutrition, Liver Unit, Avenida San Fernando y Viaducto Tlalpan, Mexico 22, Distrito Federal. The authors thank Mrs. Elizabeth -Rodrfgues Hernandez and Mrs. Edvina Robles Navarro for their secretarial help. Dr. Harold 0. Conn provided valuable comments and criticisms. Rebeca France, Ph.D., reviewed the manuscript for English usage. This investigation was supported by grants from the Consejo National de Ciencia y Tecnologia and the Academia National de Medicina, Chinoin Award (MBxico). Mycifradin and placebo tablets were kindly provided by Laboratorios Upjohn de MBxico, S.A. 0 1981 by the American Gastroenterological Association 0016-5085/81/070101-07$02.50
Unit,
The beneficial effect of lactulose in the management of patient8 with chronic and acute portal systemic encephalopathy (PSE) has been recently demonstrated in double-blind studies (1,2). In acute PSE lactulose syrup plus placebo tablet8 were as effective as neomycin tablet8 plus sorbitol syrup for the control of PSE. We and others (34 had previously demonstrated the effectiveness of lactose when administered orally to patients with chronic PSE and lactose intolerance. Among these patients theoretically lactose acts similarly to lactulose, lowering coionic pH and decreasing the intestinal time available both for production and absorption of ammonia (5). Further, in recent in vitro studies (6,7) the similar effects of lactulose and lactose on fecal ammonia generation, fecal flora, and fecal pH acidification have been demonstrated. In noncontrolled studies (8,9), lactulose enema8 seem to be effective for the control of acute PSE. Therefore, we thought that lactose enema8 may be similarly useful for the control of patients with acute PSE. The current investigation was undertaken to compare lactose enemas plus placebo tablet8 vs. starch enemas plus neomycin tablets in a double-blind, randomized study.
Methods Acute PSE was arbitrarily defined (1) as the sudden development in a cirrhotic (biopsy-proven) patient of encephalopathy precipitated by nitrogenous substances as a main factor, although patients in whom no cause of
102
URIBE ET AL.
GASTROENTEROLOGY Vol. 81, No. 1
acute PSE (endogenous) was found were also studied. None of the patients ingested analgesics, sedatives, or presented acute renal failure. Before inclusion in the study the patients must have developed within 24 h an acute episode of PSE, namely; encephalopathy of at least grade 2+ severity (2) plus two of the following abnormalities: (a) arterial ammonia levels above 120 pg% (normal ~90 pg%); (b) abnorfnal slow waves in the electroencephalogram (EEG) as blindly judged by a neurologist (GGR); and (c) time the patient needed to perform a number-connection test (10) at least double the normal range (>60 s; normal < 30 s), or patient unable to perform the test due to mental confusion or coma. Patients were excluded if they: (a) required or had ingested antibiotics; (b) presented active bleeding; (c) presented anorectal disease; (d) had a history of previous neurologic disease other than PSE; or (e) relatives refused to sign consent form. A total of 20 patients fulfilled the above criteria; however, within 24 h 2 of them developed acCte pneumonia requiring systemic antibiotics and were therefore not included in this report. To maintain the double-blind status of the study it was necessary to use a double-drug design. Eight patients received t.i.d. a 1-Lenema of 20% lactose prepared in boiled water within 24 h before its use. Siniultaneously this group received two placebo tablets t.i.d., which were identical to 0.5-g tablets of neomycin: and were kindly prepared by Laboratorios Upjohn de MBxico S.A., MBxico City. Ten patients received 1-L starch enemas (10%) plus two 0.5-g neomycin tablets t.i.d. (Mycifradin). Lactose and starch enemas were prepared and bdttled in dark-brown glass containers, which made it impossible to distinguish between starch and lactose eneinas. Lactose enemas had a pH of 4.6 f 0;17, and starch enemas a pH of 4.5 f 0.23. Similarity of pHs was not caused by the addition of any other substance. Nurses caring for the patients were instructed to shake the enema bottles strongly before their use. Starch and lactose USP purity grade were purchased from Cosmopolitan Drugs S.A., Mexico City. During the trial 5 patients of each group received initially 5%-IO%-dextrose solutions i.v.; no other medications were indicated. Improvement of 1 grade of mental status was accompanied by a increment of 20 g protein (day) to reach a maximum of 66 g/day. The following parameters (11) were assessed before entry to the trial and daily thereafter.
Mental
State
As described by Conn et al. (l), mental graded from normal to coma (0 to 4+).
Number-Connection
state was
Test
This test (10) is the time in seconds required for the patients to connect 25 circled numbers. Td avoid the learning effect, four different versions of the number-connection test (NCT) were used. Versions of the NCT designated A, B, C, and D were given to 40 individuals who
served as controls and who had a similar educational level as the patients. The exhibited control score for version A ranged from 11 s to 25 s, mean 16s; for version B, the range was from 11 to 27 s, mean 19 s; for version C, from 10 to 31 s, mean 21 s; and for version D, from 12 to 25 s, mean 20 s. As the NCT is the time expressed in secorids required to connect 25 numbers, it was arbitrarily graded as followd: grade 0, <3b s; grade l+, 31-60 s; grade 2+, 61-100 s; grade 3+, 100-200 s; and grade 4+, <206 s; or patients unable to perform the test.
Asterixis Presence of asterixis was graded as follows: grade 0, no flapping motion; grade l+, rare, 5 flaps/min; grade 2+, occasional irregular flaps, 6-lO/min; grade 3+, freqtient flaps, ll-20/min; grade 4+, almost continuous flapping motions, or patient in coma and unable to maintain wrist dorsiflexion.
Electroencephalograms Electroencephalograms were iaken by a model Van Gogh E PB 8 that uses standard lo-20 electrode placements (Ahrenil Co., The Netherlands). Electrdencephalograms were read blindly and graded by one of us (G.G.R.), who was unaware of the kind of medication given to the patients. Tracings were assessed semiquantitatively on a 0 to 4+ scale; grade 0, was normal alpha rhythm; grade l+, theta activity 7-8 cycles/s; grade 2+, 5-7 cycles/s: grade 3+, 3-5 cycles/s; grade 4+, <3 cycles/s, or delta rhythm.
Blood Ammonia
Concentration
Fasting arterial blood ammonia levels were measured (12) by using a Beckman microtitrator (model 153; Beckman Instruments, Fullerton, Calif.). It was arbitrarily graded as follows: grade 0, < 90 pg; grade l+, 91-120;. grade 2+, 121-150; grade 3+, 151-180; and grade 4+ >180 pg% (normal: 60-90 pg/lOO ml).
Portal Systemic Encephalopathy
Index
Each parameter was arbitrarily (2) weighed according to its importance. Thus, mental state was assigned a factor of 3 and each of the others a factor of 1. The overall maximum PSE score being 28 (4 X 3 = 12; 4 X 4 = 16; thus, 16 + 12 = 28). The total weight of the score was called the @SE sum. The ratio PSE sum/maximum PSE gives the PSE index that arbitrarily reflects the changes in all parameters produced during the study.
Fecal pH All stools were collected immediately and pH determined by pH paper. (Phydrion, Micro Essential Laboratory, Brooklyn N-Y.). The paper has been shown to be reliable, particularly below 6.0 (13). The number of stools were also recorded.
luly 1981
Table
1.
LACTOSE IN ACUTE PSE
Similarities
of the Groups Studied
Age(YV Sex (M/F) Type of cirrhosis: alcoholic/nonalcoholic Surgical portacaval shunt Cause of acute PSEb Gastrointestinal bleeding Protein diet Diuretics Unknown Time with PSE clinical” signs (yr) Jaundice Hepatomegaly Ascitis SGOT, IU/dl”*” SGPT, IIJ/dl”,d Conjugated bilirubin (mgW Unconjugated bilirubin (mg%) Alkaline phosphatase (U/W’ Prothrombin time (s)~
Starch enemas plus neomycin tablets (n = 20)
Lactose enemas plus placebo tablets (n = 8)
p-Value
55 f 9 2/8
51+ 11 4/4
NS NS
5/5 3
6/2 3
NS/NS
3 1 2 4
2 1 2 3
NS NS NS NS
1.7 f 1.3 5 4 3 68 f 70 44 f 40
1.5 f 1.4 5 3 2 67 2 59 36f26
NS NS NS NS NS NS
3.4 zt 5.6
5.3 f 5.9
NS
1.7 f 1.7
3.2 f 4.0
NS
84 * 22 16.7 f 2.4
77 + 49 17.3 f 2.8
NS NS
Patients were continued on the regimen until they achieved and maintained one or more grades of mental state improvement for at least 48 h, then oral treatment was administered (lactose or neomycin); at this time the study was considered concluded. In patients without mental state response after 3-4 days of treatment, the study was considered concluded, and the last two sets of data were used for calculations.
Statistical
tients in the Gastrointestinal Unit of the Hospital General, Centro Medico National IMSS, under the supervision of one of us (M. Dehesa). The nature, purpose, and hazards of this study were explained to the patients (when possible) and their closest relatives; then informed consent was obtained. The protocal of this study has been approved by the Clinical Investigation Committee of the Instituto National de la Nutricibn, Mexico City.
Results
NS
a Mean value k SD. b PSE = portal systemic encephalopathy. c SGOT = serum glutamic oxalacetic transaminase. d SGPT = serum piruvic transaminase.
Patients were similar with regard to clinicalbiochemical characteristics, kind and duration of cirrhosis, frequency of portacaval shunts, etiology, severity of PSE, duration of therapy, and protein ingestion (Tables 1 and 2). Clinical Efficacy Clinical-biochemical improvement was observed after lactose enemas in 7 of 8 patients (87%) and in 7 of 10 patients (70%) treated with starch enemas plus neomycin. This difference was not statistically significant. Mental State After both treatments 5 patients were discharged with normal mental state (grade 0). Both treatments significantly improved mental state from 2.7 + 0.8 to 1.1 +- 1.4 (p c 0.05) after starch enemas plus neomycin, and from 2.8 f 0.8 to 0.87 A 1.4 after lactose enemas plus placebo tablets (p < 0.025; Figure 1).
Table 2.
Portal Systemic Encephalopathy the Groups”
PSE parameters and laboratory tests. Comparisons were done with individual data before treatments vs. the mean of the last two sets of data obtained with each patient. The &i-square technique (14) was used for analysis of patients clinical characteristics (Table 1) during the study. Clinical-biochemical response was arbitrarily defined as the improvement of 1 grade in mental state, a reduction of
30 s in the time to perform the NCTs and ammonia reduction of at least 50 pg%. Sixteen patients were studied in the Special Care Unit of the Instituto National de la Nutrici6n under the supervision of one of us (M. Uribe), and the remaining 2 pa-
Parameters
of
Starch Lactose enemas plus enemas plus neomycin placebo (n = 20) p-Value (n = 6)
Analysis
All analyses were performed without knowledge of the nature of the treatment administered. Data were analysed by using the paired Student’s t-test for analysis of
103
Mental statebpc Number-connection test (s) Asterixisbsc Electroencephalogramb,’ Arterial ammonia @g%) Portal systemic encephalopathy index” Clinical biochemical efficacyd No. stools per dayd Protein intake (g/day)d Duration of treatment (day)
2.7 f 0.82
2.8 rt 0.6
NS
24OfO 3.4 f 1.2 3.7 f 0.6 191+ 124
237 f 6 3.8 * 0.4 3.6f0.6 218 f 14
NS
0.72 f 0.20 7 of 10 (70%) 3.9 f 0.7 25 f 9 4.0 f 0.9
0.80 f 0.13 7 of a (87%) 4.5 f 1.0 24 f 12 4.1 i 0.8
n Values are expressed as mean f SD. b Graded on l-4+ c At the time of randomization. d After treatment.
NS NS NS NS NS NS NS scale.
104
GASTROENTEROLOGY
URIBE ET AL.
m BEFORE ?? AFTER ?? BEFORE ?? AFTER
3t MENTAL STATE GRADE
2+ It 0
Asterixis
NEOMVCIN NEOMYCIN
The mean grade of asterixis improved (p < 0.05and p < 0.005 after starch enemas-neomycin and lactose enemas, respectively). Asterixis disappeared in 5 patients treated with starch enemasneomycin and in 4 patients treated with lactose enemas plus placebo tablets (Figure 1).
LACTOSE LACTOSE
A
VI
B <
0.05
c
“8
D <
0.025
A
VI
E
(0.0s
C
VI
D
CO.005
4t ASTERIXIS GRADE
‘+ z+ It 0
Electroencephalograms Both treatments significantly improved EEG Figure 1);in 3 abnormalities (p < 0.005 and p < 0.005; cases, starch enemas-neomycin restored the EEG features to normal (alpha rhythm, grade 0,). The same occurred in 3 patients treated with lactose enemas placebo tablets.
3t EEG GRADE
Vol. 81,No.1
” It 0 A
c
I)
A
VI
B
CO.005
c
“l
0
<0.005
D
Figure 1. Effect of starch enemas-neomycin and lactose enemasplacebo on mental state, asterixis, and EEGs. Bars represent mean f SD. Only significant differences are shown.
In all but 1 of the patients who grade of mental state improvement within the first 24 h (Figure 2). There ences in days of follow-up among studied (Table 2). Number-Connection
Blood Ammonia
Rapid reductions of blood ammonia levels were noticed regardless of the treatment used. The changes were of significant value (p < 0.05and p < 0.025 after starch enemas-neomycin and lactose enemas-placebo tablets, respectively.) Blood ammonia levels were restored to normal value in 3 patients on starch enemas-neomycin and in 4 patients on lactose enemas (Figure 3).
responded, one was noticeable were no differthe two groups
Test Fecal pH
After lactose enemas, NCT improved significantly (p < 0.02). After starch enemas-neomycin NCT improved, but did not reach a significant value (Figure 3). MENTAL
Stool pH decreased from 6.0 f 0.6 to 5.6f 0.75 in the starch enemas-neomycin group (Figure 9, but this change was not significant. In contrast, fecal pH STATE
EVOLUTION WITH
0
I
Levels
a
a DAYS
Figure 2. Changes in mental state as blindly represents 1 patient.
evaluated
4
6
AFTER after starch
0 1rREATYEN1
I
enemas-neomycin
NCOYYCIN
0
and lactose
1
4
enemas-placebo.
5
Each symbol
July 1981
LACTOSE IN ACUTE PSE
??BEFORE
NEOMYCIN
AFTER ??
NEOMYCIN
i&
BEFORE
&J AFTER 200
LACTOSE LACTOSE
105
served in liver function tests or other laboratory parameters. Within 1 mo after the end of the study one patient of each group died due to liver failure and to esophageal bleeding.
--
AMMONIA CONCENTRATIONS PQ%
Discussion
100 A
Y,
B
< 0.05
C “I
D
-z 0.025
250 NUMBER
200
CONNECTION
,SQ
TEST.
,oo
SEC.
10
c A
B
c
“8
D<0.02
0
Figure 3. Effect of starch enemas-neomycin and lactose enemasplacebo on ammonia concentrations and number-connection test. Only significant differences are shown.
significantly decreased in the lactose enemas treated group from 6.0f 0.6to 4.6f 1.0(p < 0.05). Stool pH was acidified after treatment in 5 of 10 patients on starch enemas-neomycin and in 7 of 6 patients on lactose enemas-placebo tablets. Portal Systemic
Encephalopathy
Jndex
After treatment with starch enemas-neomycin PSE index improved from 0.72f 0.2to 0.38f 0.20(p < 0.01) (Figure 5).The same phenomenon occurred after treatment with lactose enemas-placebo, in this case PSE index decreased from 0.8f 0.1to 0.32f 0.3(p < 0.005). Side Effects No clinical side effects were noticed with the treatments used. No significant changes were ob-
LACTOSE
NEOMYCIN
Lactulose enemas have been used (15)in 72 acute episodes of PSE and resulted effective in 60 of them (83%).Similarities between the effects of lactose and lactulose on fecal in vitro incubations (6,7) prompted us to explore the use of lactose enemas in acute PSE. Not surprisingly, the results of our double-blind, randomized study resembled the results obtained by Atterbury et al. (1) with lactulose. Indeed, lactose enemas were as effective as starch enemas-neomycin in the control of acute PSE. There was a similar, highly significant improvement in each of the clinical and laboratory components of the PSE syndrome. This improvement was also demonstrable by a similar change in the PSE index after both treatments. Lactose enemas significantly reduced stool pH whereas starch-neomycin did not. When administered directly into the colon, lactose degrades rapidly, thus saving time by eliminating the long, tortuous trip through the gut. Thus, the rectal route may be advantageous over the oral administration in this particular population of severely ill patients. It is possible that lactose-like lactulose-may be rapidly metabolized in the colon. Bond and Levitt (16) infused lactulose into the cecum in a group of patients and observed the appearance of hydrogen gas in expired air within 4-5 min. If the same occurs after lactose enemas, this would explain the rapid clinical response observed. Both treatments significantly improved mental state and reduced blood ammonia levels. A combination of factors may explain these results. First, administration of lactose enemas leads to colonic pH acidification that in turn may reduce the diffusion of ammonia from the colon to the systemic circulation (17). Second, as occurred with lactulose, the presence of a fermentable carbohydrate like lactose may
m 1.0
PSE INDEX
BEFORE
Figure
AFTER
BEFORE
4. Changes in stool pH after treatment with enemas-neomycin and lactose enemas-placebo.
AFTER
starch
1,
BEFORE
NEOMYCIN
? ?AFTERNEOMYCIN
T
0.8
?? BEFORE
LACTOSE
0.6
&j
LACTOSE
AFTER
0.4
A
B
c
D
Figure 5. Effect of starch enemas-neomycin placebo on PSE index.
and lactose enemas-
106
GASTROENTEROLOGY
LJRIBE ET AL.
decrease blood ammonia by providing an energy source and thus spare the metabolism of exogenous aminated compounds with subsequent decrease in the generation of ammonia by colonic bacteria. Third, 20% lactose enemas have an osmolality of 364 mosmol/L, whereas 10% starch enemas have very little osmolality (2 mosmol/L). The same applies to hypertonic lactulose enemas used by others. Bircher et al. (18)have shown that such an effect (osmotic) is important in the catharsis of lactulose. The beneficial effects obtained with starch enemas-neomycin may also be explained by a combination of factors. Theoretically, starch may be metabolized to some extent by colonic bacteria. However, the small changes on fecal pH observed after this treatment makes this possibility unlikely. It has been demonstrated however, that administration of neomycin reduces the output of ammonia by the colon (19). The acidity of both enemas (starch or lactose) raises the possibility that, at least for an instant, coionic pH was acidified, and therefore this may decrease ammonia absorption. However, this effect was maintained only after lactose enemas. Thus, sustained, pH-dependent effect can only be implicated during lactose treatment. Finally, it cannot be ruled out that the improvement observed after both treatments could have been to a significant extent the consequence of the cleansing effect of both enemas. Nonetheless sterilization induced by neomycin, or colonic acidification induced by lactose should not be ignored. In fact, colonic ammonia output is more reduced by a combination of cleansing-sterilization than compared with only one of these methods used separately (19). To investigate the isolated effect of lactose on PSE a study should be done comparing lactose enemas with tap-water enemas. Such a study was planned in our unit, but for ethical reasons our Clinical Investigation Committee did not approve its performance. Lactose enemas are an effective alternative to neomycin, cathartics, or lactulose in the treatment of acute PSE. When deciding whether or not to use lactose enemas, it is not mandatory that patients be lactose intolerant. Lactose enemas are particularly useful in patients with ilei in whom the oral administration of lactose or lactulose is not feasible. Although the population studied is small, lactose enemas were free of side effects. Also, the enemas are inexpensive and easy to prepare (in Mexico, 1 kg of lactose costs $1.50), an important advantage in developing countries such as Mexico.
Vol. 81, No. 1
References 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. 2. Conn 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. Uribe M, Marquez MA, Garcfa Ramos G, et al. Treatment of chronic portal systemic encephalopathy with lactose in lactase defficient patients. Dig Dis Sci 1980;25:92.4-8. 4. Welch JD, Cassidy D, Prigatano GP, et al. Chronic hepatic encephalopathy treated with oral lactose in a patient with lactose malabsorption. N Engl J Med 1974;291:240-1. 5. Hoyumpa AM, Desmond PV, Avant PR, et al. Hepatic encephalopathy. Gastroenterology 1977;76:184-95. 6. Lewis H, Rojas S, Uribe M, et al. Similar effect of lactose and lactulose on in vitro fecal ammonia generation and bacterial flora (abstr). Gastroenterology 1986;78:1206. 7. Vince AJ, Burridge SM. Ammonia production by intestinal bacteria: the effects of lactose, lactulose and glucose. J Med Microbial 1980;13:177-92. 8. Imler M, Kurtz D, Bockel R, et al. Etude comparative du traitment de 1 encephalopathie portocave par le lactulose, les batiles lactiques et les antibiotiques. Therapeutique 1971; 47:237-48. 9. Kersh ES, Rifkin H. Lactulose enemas. Ann Intern Med 1973;78:81-4. 10.Conn HO. Trailmaking and number connection tests in the assessment of mental state in portal systemic encephalopathy. Dig Dis Sci 1977;22:541-50. 11.Uribe M, Farca A, Marquez MA, et al. Treatment of chronic portal systemic encephalopathy with bromocriptine a doubleblind controlled trial. Gastroenterology 1979;76:1347-51. 12. MC Dermont WV, Adams PD, Riddell AG. Ammonia levels in blood and cerebrospinal fluid. Proc Sot Exp Biol 1955;88:3802. 13. Conn HO, Lieberthal MM. Measurement of stool pH. In: Conn HO, Lieberthal MM, eds. The hepatic coma syndromes and lactulose. Chapt 11. Baltimore: Williams & Wilkins, 1979:295305. 14.Rickerms A, Todd HN. Introducci6n a la estadistica. 2nd ed. Barcelona, EspaAa: Editado por la Compaiiia Editorial Continental SA, 1974:90-3. 15. Conn HO, Lieberthal MM. Lactulose enemas. In: Conn HO, Lieberthal MM, eds. The hepatic coma syndromes and lactulose. Chap 13. Baltimore: Williams & Wilkins, 1979:317-22. 16.Bond JH, Jr, Levitt MD. Investigation of small bowel transit time in man utilizing pulmonary hydrogen (H,) measurements J Lab Clin Med 1979;85:546-55. 17.Price JB, Jr, Sawada M, Voorhees AB, Jr. Clinical significance of intraluminal pH in intestinal ammonia transport. Am J Surg 1970;119:595-8. 18. Pertsiounis S, Egger G, Bircher J. Radiological “quantitation” of osmotic effect. Baden bei Wein, Austria: Symposium on Lactulose, 1969. 19. Wolpert E, Phillips SF, Summerskill WHJ. Ammonia production in the human colon. Effects of cleansing, neomycin and acetohydroxamic acid. N Engl J Med 1970;283:159-64. 1.
LIVER
PHYSIOLOGY
AND
1981;81:101-6
DISEASE
Lactose Enemas Plus Placebo Tablets vs. Neomycin Tablets Plus Starch Enemas in Acute Portal Systemic Encephalopathy A Double-Blind Randomized
Controlled Study
MISAEL URIBE, JESIjS MORENO BkRTHIER, HILEL LEWIS, JUAN M. MATA, JOSE G. SIERRA, GUILLERMO GARCfA-RAMOS, JAVIER RAMfREZ ACOSTA, and MARGARITA DEHESA, with the Technical Assistance of ELBA GALVAN and SARA ALVAREZ The Liver Unit, Instituto National de la Nutricibn, and the Gastrointestinal Hospital General IMSS MBxico City, MBxico
A randomized, double-blind comparison of lactose enemas plus placebo tablets vs. starch enemas plus neomycin tablets was performed on 18 patients with acute portal systemic encephalopathy. Ten patients received starch enemas (10%;2000 ml t.i.d.) plus neomycin tablets and 8 patients received lactose enemas (20%;1000 ml Cd.) plus placebo tablets. A significant mental state improvement was demonstrated in the group of patients treated with starch enemas-neomycin tablets (p c 0.05) and in the group of patients treated with lactose enemas-placebo tabBoth treatments significantly imlets (p < 0.025). proved the frequency of asterixis, ammonia blood levels, and electroencephalograms. In addition, patients treated with lactose enemas showed significant improvement in number-connection test times (p < O.O2), and their stools showed a mQre acid pH (p c 0.05). No side effects were evident with either treatment. Lactose enemas are a safe and effective treatment for acute portal systemic encephalopathy. Received September 3, 1980. Accepted February 7, 1981. Address requests for reprints to: Dr. Misael Uribe, Instituto Nacional de la Nutrition, Liver Unit, Avenida San Fernando y Viaducto Tlalpan, Mexico 22, Distrito Federal. The authors thank Mrs. Elizabeth -Rodrfgues Hernandez and Mrs. Edvina Robles Navarro for their secretarial help. Dr. Harold 0. Conn provided valuable comments and criticisms. Rebeca France, Ph.D., reviewed the manuscript for English usage. This investigation was supported by grants from the Consejo National de Ciencia y Tecnologia and the Academia National de Medicina, Chinoin Award (MBxico). Mycifradin and placebo tablets were kindly provided by Laboratorios Upjohn de MBxico, S.A. 0 1981 by the American Gastroenterological Association 0016-5085/81/070101-07$02.50
Unit,
The beneficial effect of lactulose in the management of patient8 with chronic and acute portal systemic encephalopathy (PSE) has been recently demonstrated in double-blind studies (1,2). In acute PSE lactulose syrup plus placebo tablet8 were as effective as neomycin tablet8 plus sorbitol syrup for the control of PSE. We and others (34 had previously demonstrated the effectiveness of lactose when administered orally to patients with chronic PSE and lactose intolerance. Among these patients theoretically lactose acts similarly to lactulose, lowering coionic pH and decreasing the intestinal time available both for production and absorption of ammonia (5). Further, in recent in vitro studies (6,7) the similar effects of lactulose and lactose on fecal ammonia generation, fecal flora, and fecal pH acidification have been demonstrated. In noncontrolled studies (8,9), lactulose enema8 seem to be effective for the control of acute PSE. Therefore, we thought that lactose enema8 may be similarly useful for the control of patients with acute PSE. The current investigation was undertaken to compare lactose enemas plus placebo tablet8 vs. starch enemas plus neomycin tablets in a double-blind, randomized study.
Methods Acute PSE was arbitrarily defined (1) as the sudden development in a cirrhotic (biopsy-proven) patient of encephalopathy precipitated by nitrogenous substances as a main factor, although patients in whom no cause of
102
URIBE ET AL.
GASTROENTEROLOGY Vol. 81, No. 1
acute PSE (endogenous) was found were also studied. None of the patients ingested analgesics, sedatives, or presented acute renal failure. Before inclusion in the study the patients must have developed within 24 h an acute episode of PSE, namely; encephalopathy of at least grade 2+ severity (2) plus two of the following abnormalities: (a) arterial ammonia levels above 120 pg% (normal ~90 pg%); (b) abnorfnal slow waves in the electroencephalogram (EEG) as blindly judged by a neurologist (GGR); and (c) time the patient needed to perform a number-connection test (10) at least double the normal range (>60 s; normal < 30 s), or patient unable to perform the test due to mental confusion or coma. Patients were excluded if they: (a) required or had ingested antibiotics; (b) presented active bleeding; (c) presented anorectal disease; (d) had a history of previous neurologic disease other than PSE; or (e) relatives refused to sign consent form. A total of 20 patients fulfilled the above criteria; however, within 24 h 2 of them developed acCte pneumonia requiring systemic antibiotics and were therefore not included in this report. To maintain the double-blind status of the study it was necessary to use a double-drug design. Eight patients received t.i.d. a 1-Lenema of 20% lactose prepared in boiled water within 24 h before its use. Siniultaneously this group received two placebo tablets t.i.d., which were identical to 0.5-g tablets of neomycin: and were kindly prepared by Laboratorios Upjohn de MBxico S.A., MBxico City. Ten patients received 1-L starch enemas (10%) plus two 0.5-g neomycin tablets t.i.d. (Mycifradin). Lactose and starch enemas were prepared and bdttled in dark-brown glass containers, which made it impossible to distinguish between starch and lactose eneinas. Lactose enemas had a pH of 4.6 f 0;17, and starch enemas a pH of 4.5 f 0.23. Similarity of pHs was not caused by the addition of any other substance. Nurses caring for the patients were instructed to shake the enema bottles strongly before their use. Starch and lactose USP purity grade were purchased from Cosmopolitan Drugs S.A., Mexico City. During the trial 5 patients of each group received initially 5%-IO%-dextrose solutions i.v.; no other medications were indicated. Improvement of 1 grade of mental status was accompanied by a increment of 20 g protein (day) to reach a maximum of 66 g/day. The following parameters (11) were assessed before entry to the trial and daily thereafter.
Mental
State
As described by Conn et al. (l), mental graded from normal to coma (0 to 4+).
Number-Connection
state was
Test
This test (10) is the time in seconds required for the patients to connect 25 circled numbers. Td avoid the learning effect, four different versions of the number-connection test (NCT) were used. Versions of the NCT designated A, B, C, and D were given to 40 individuals who
served as controls and who had a similar educational level as the patients. The exhibited control score for version A ranged from 11 s to 25 s, mean 16s; for version B, the range was from 11 to 27 s, mean 19 s; for version C, from 10 to 31 s, mean 21 s; and for version D, from 12 to 25 s, mean 20 s. As the NCT is the time expressed in secorids required to connect 25 numbers, it was arbitrarily graded as followd: grade 0, <3b s; grade l+, 31-60 s; grade 2+, 61-100 s; grade 3+, 100-200 s; and grade 4+, <206 s; or patients unable to perform the test.
Asterixis Presence of asterixis was graded as follows: grade 0, no flapping motion; grade l+, rare, 5 flaps/min; grade 2+, occasional irregular flaps, 6-lO/min; grade 3+, freqtient flaps, ll-20/min; grade 4+, almost continuous flapping motions, or patient in coma and unable to maintain wrist dorsiflexion.
Electroencephalograms Electroencephalograms were iaken by a model Van Gogh E PB 8 that uses standard lo-20 electrode placements (Ahrenil Co., The Netherlands). Electrdencephalograms were read blindly and graded by one of us (G.G.R.), who was unaware of the kind of medication given to the patients. Tracings were assessed semiquantitatively on a 0 to 4+ scale; grade 0, was normal alpha rhythm; grade l+, theta activity 7-8 cycles/s; grade 2+, 5-7 cycles/s: grade 3+, 3-5 cycles/s; grade 4+, <3 cycles/s, or delta rhythm.
Blood Ammonia
Concentration
Fasting arterial blood ammonia levels were measured (12) by using a Beckman microtitrator (model 153; Beckman Instruments, Fullerton, Calif.). It was arbitrarily graded as follows: grade 0, < 90 pg; grade l+, 91-120;. grade 2+, 121-150; grade 3+, 151-180; and grade 4+ >180 pg% (normal: 60-90 pg/lOO ml).
Portal Systemic Encephalopathy
Index
Each parameter was arbitrarily (2) weighed according to its importance. Thus, mental state was assigned a factor of 3 and each of the others a factor of 1. The overall maximum PSE score being 28 (4 X 3 = 12; 4 X 4 = 16; thus, 16 + 12 = 28). The total weight of the score was called the @SE sum. The ratio PSE sum/maximum PSE gives the PSE index that arbitrarily reflects the changes in all parameters produced during the study.
Fecal pH All stools were collected immediately and pH determined by pH paper. (Phydrion, Micro Essential Laboratory, Brooklyn N-Y.). The paper has been shown to be reliable, particularly below 6.0 (13). The number of stools were also recorded.
luly 1981
Table
1.
LACTOSE IN ACUTE PSE
Similarities
of the Groups Studied
Age(YV Sex (M/F) Type of cirrhosis: alcoholic/nonalcoholic Surgical portacaval shunt Cause of acute PSEb Gastrointestinal bleeding Protein diet Diuretics Unknown Time with PSE clinical” signs (yr) Jaundice Hepatomegaly Ascitis SGOT, IU/dl”*” SGPT, IIJ/dl”,d Conjugated bilirubin (mgW Unconjugated bilirubin (mg%) Alkaline phosphatase (U/W’ Prothrombin time (s)~
Starch enemas plus neomycin tablets (n = 20)
Lactose enemas plus placebo tablets (n = 8)
p-Value
55 f 9 2/8
51+ 11 4/4
NS NS
5/5 3
6/2 3
NS/NS
3 1 2 4
2 1 2 3
NS NS NS NS
1.7 f 1.3 5 4 3 68 f 70 44 f 40
1.5 f 1.4 5 3 2 67 2 59 36f26
NS NS NS NS NS NS
3.4 zt 5.6
5.3 f 5.9
NS
1.7 f 1.7
3.2 f 4.0
NS
84 * 22 16.7 f 2.4
77 + 49 17.3 f 2.8
NS NS
Patients were continued on the regimen until they achieved and maintained one or more grades of mental state improvement for at least 48 h, then oral treatment was administered (lactose or neomycin); at this time the study was considered concluded. In patients without mental state response after 3-4 days of treatment, the study was considered concluded, and the last two sets of data were used for calculations.
Statistical
tients in the Gastrointestinal Unit of the Hospital General, Centro Medico National IMSS, under the supervision of one of us (M. Dehesa). The nature, purpose, and hazards of this study were explained to the patients (when possible) and their closest relatives; then informed consent was obtained. The protocal of this study has been approved by the Clinical Investigation Committee of the Instituto National de la Nutricibn, Mexico City.
Results
NS
a Mean value k SD. b PSE = portal systemic encephalopathy. c SGOT = serum glutamic oxalacetic transaminase. d SGPT = serum piruvic transaminase.
Patients were similar with regard to clinicalbiochemical characteristics, kind and duration of cirrhosis, frequency of portacaval shunts, etiology, severity of PSE, duration of therapy, and protein ingestion (Tables 1 and 2). Clinical Efficacy Clinical-biochemical improvement was observed after lactose enemas in 7 of 8 patients (87%) and in 7 of 10 patients (70%) treated with starch enemas plus neomycin. This difference was not statistically significant. Mental State After both treatments 5 patients were discharged with normal mental state (grade 0). Both treatments significantly improved mental state from 2.7 + 0.8 to 1.1 +- 1.4 (p c 0.05) after starch enemas plus neomycin, and from 2.8 f 0.8 to 0.87 A 1.4 after lactose enemas plus placebo tablets (p < 0.025; Figure 1).
Table 2.
Portal Systemic Encephalopathy the Groups”
PSE parameters and laboratory tests. Comparisons were done with individual data before treatments vs. the mean of the last two sets of data obtained with each patient. The &i-square technique (14) was used for analysis of patients clinical characteristics (Table 1) during the study. Clinical-biochemical response was arbitrarily defined as the improvement of 1 grade in mental state, a reduction of
30 s in the time to perform the NCTs and ammonia reduction of at least 50 pg%. Sixteen patients were studied in the Special Care Unit of the Instituto National de la Nutrici6n under the supervision of one of us (M. Uribe), and the remaining 2 pa-
Parameters
of
Starch Lactose enemas plus enemas plus neomycin placebo (n = 20) p-Value (n = 6)
Analysis
All analyses were performed without knowledge of the nature of the treatment administered. Data were analysed by using the paired Student’s t-test for analysis of
103
Mental statebpc Number-connection test (s) Asterixisbsc Electroencephalogramb,’ Arterial ammonia @g%) Portal systemic encephalopathy index” Clinical biochemical efficacyd No. stools per dayd Protein intake (g/day)d Duration of treatment (day)
2.7 f 0.82
2.8 rt 0.6
NS
24OfO 3.4 f 1.2 3.7 f 0.6 191+ 124
237 f 6 3.8 * 0.4 3.6f0.6 218 f 14
NS
0.72 f 0.20 7 of 10 (70%) 3.9 f 0.7 25 f 9 4.0 f 0.9
0.80 f 0.13 7 of a (87%) 4.5 f 1.0 24 f 12 4.1 i 0.8
n Values are expressed as mean f SD. b Graded on l-4+ c At the time of randomization. d After treatment.
NS NS NS NS NS NS NS scale.
104
GASTROENTEROLOGY
URIBE ET AL.
m BEFORE ?? AFTER ?? BEFORE ?? AFTER
3t MENTAL STATE GRADE
2+ It 0
Asterixis
NEOMVCIN NEOMYCIN
The mean grade of asterixis improved (p < 0.05and p < 0.005 after starch enemas-neomycin and lactose enemas, respectively). Asterixis disappeared in 5 patients treated with starch enemasneomycin and in 4 patients treated with lactose enemas plus placebo tablets (Figure 1).
LACTOSE LACTOSE
A
VI
B <
0.05
c
“8
D <
0.025
A
VI
E
(0.0s
C
VI
D
CO.005
4t ASTERIXIS GRADE
‘+ z+ It 0
Electroencephalograms Both treatments significantly improved EEG Figure 1);in 3 abnormalities (p < 0.005 and p < 0.005; cases, starch enemas-neomycin restored the EEG features to normal (alpha rhythm, grade 0,). The same occurred in 3 patients treated with lactose enemas placebo tablets.
3t EEG GRADE
Vol. 81,No.1
” It 0 A
c
I)
A
VI
B
CO.005
c
“l
0
<0.005
D
Figure 1. Effect of starch enemas-neomycin and lactose enemasplacebo on mental state, asterixis, and EEGs. Bars represent mean f SD. Only significant differences are shown.
In all but 1 of the patients who grade of mental state improvement within the first 24 h (Figure 2). There ences in days of follow-up among studied (Table 2). Number-Connection
Blood Ammonia
Rapid reductions of blood ammonia levels were noticed regardless of the treatment used. The changes were of significant value (p < 0.05and p < 0.025 after starch enemas-neomycin and lactose enemas-placebo tablets, respectively.) Blood ammonia levels were restored to normal value in 3 patients on starch enemas-neomycin and in 4 patients on lactose enemas (Figure 3).
responded, one was noticeable were no differthe two groups
Test Fecal pH
After lactose enemas, NCT improved significantly (p < 0.02). After starch enemas-neomycin NCT improved, but did not reach a significant value (Figure 3). MENTAL
Stool pH decreased from 6.0 f 0.6 to 5.6f 0.75 in the starch enemas-neomycin group (Figure 9, but this change was not significant. In contrast, fecal pH STATE
EVOLUTION WITH
0
I
Levels
a
a DAYS
Figure 2. Changes in mental state as blindly represents 1 patient.
evaluated
4
6
AFTER after starch
0 1rREATYEN1
I
enemas-neomycin
NCOYYCIN
0
and lactose
1
4
enemas-placebo.
5
Each symbol
July 1981
LACTOSE IN ACUTE PSE
??BEFORE
NEOMYCIN
AFTER ??
NEOMYCIN
i&
BEFORE
&J AFTER 200
LACTOSE LACTOSE
105
served in liver function tests or other laboratory parameters. Within 1 mo after the end of the study one patient of each group died due to liver failure and to esophageal bleeding.
--
AMMONIA CONCENTRATIONS PQ%
Discussion
100 A
Y,
B
< 0.05
C “I
D
-z 0.025
250 NUMBER
200
CONNECTION
,SQ
TEST.
,oo
SEC.
10
c A
B
c
“8
D<0.02
0
Figure 3. Effect of starch enemas-neomycin and lactose enemasplacebo on ammonia concentrations and number-connection test. Only significant differences are shown.
significantly decreased in the lactose enemas treated group from 6.0f 0.6to 4.6f 1.0(p < 0.05). Stool pH was acidified after treatment in 5 of 10 patients on starch enemas-neomycin and in 7 of 6 patients on lactose enemas-placebo tablets. Portal Systemic
Encephalopathy
Jndex
After treatment with starch enemas-neomycin PSE index improved from 0.72f 0.2to 0.38f 0.20(p < 0.01) (Figure 5).The same phenomenon occurred after treatment with lactose enemas-placebo, in this case PSE index decreased from 0.8f 0.1to 0.32f 0.3(p < 0.005). Side Effects No clinical side effects were noticed with the treatments used. No significant changes were ob-
LACTOSE
NEOMYCIN
Lactulose enemas have been used (15)in 72 acute episodes of PSE and resulted effective in 60 of them (83%).Similarities between the effects of lactose and lactulose on fecal in vitro incubations (6,7) prompted us to explore the use of lactose enemas in acute PSE. Not surprisingly, the results of our double-blind, randomized study resembled the results obtained by Atterbury et al. (1) with lactulose. Indeed, lactose enemas were as effective as starch enemas-neomycin in the control of acute PSE. There was a similar, highly significant improvement in each of the clinical and laboratory components of the PSE syndrome. This improvement was also demonstrable by a similar change in the PSE index after both treatments. Lactose enemas significantly reduced stool pH whereas starch-neomycin did not. When administered directly into the colon, lactose degrades rapidly, thus saving time by eliminating the long, tortuous trip through the gut. Thus, the rectal route may be advantageous over the oral administration in this particular population of severely ill patients. It is possible that lactose-like lactulose-may be rapidly metabolized in the colon. Bond and Levitt (16) infused lactulose into the cecum in a group of patients and observed the appearance of hydrogen gas in expired air within 4-5 min. If the same occurs after lactose enemas, this would explain the rapid clinical response observed. Both treatments significantly improved mental state and reduced blood ammonia levels. A combination of factors may explain these results. First, administration of lactose enemas leads to colonic pH acidification that in turn may reduce the diffusion of ammonia from the colon to the systemic circulation (17). Second, as occurred with lactulose, the presence of a fermentable carbohydrate like lactose may
m 1.0
PSE INDEX
BEFORE
Figure
AFTER
BEFORE
4. Changes in stool pH after treatment with enemas-neomycin and lactose enemas-placebo.
AFTER
starch
1,
BEFORE
NEOMYCIN
? ?AFTERNEOMYCIN
T
0.8
?? BEFORE
LACTOSE
0.6
&j
LACTOSE
AFTER
0.4
A
B
c
D
Figure 5. Effect of starch enemas-neomycin placebo on PSE index.
and lactose enemas-
106
GASTROENTEROLOGY
LJRIBE ET AL.
decrease blood ammonia by providing an energy source and thus spare the metabolism of exogenous aminated compounds with subsequent decrease in the generation of ammonia by colonic bacteria. Third, 20% lactose enemas have an osmolality of 364 mosmol/L, whereas 10% starch enemas have very little osmolality (2 mosmol/L). The same applies to hypertonic lactulose enemas used by others. Bircher et al. (18)have shown that such an effect (osmotic) is important in the catharsis of lactulose. The beneficial effects obtained with starch enemas-neomycin may also be explained by a combination of factors. Theoretically, starch may be metabolized to some extent by colonic bacteria. However, the small changes on fecal pH observed after this treatment makes this possibility unlikely. It has been demonstrated however, that administration of neomycin reduces the output of ammonia by the colon (19). The acidity of both enemas (starch or lactose) raises the possibility that, at least for an instant, coionic pH was acidified, and therefore this may decrease ammonia absorption. However, this effect was maintained only after lactose enemas. Thus, sustained, pH-dependent effect can only be implicated during lactose treatment. Finally, it cannot be ruled out that the improvement observed after both treatments could have been to a significant extent the consequence of the cleansing effect of both enemas. Nonetheless sterilization induced by neomycin, or colonic acidification induced by lactose should not be ignored. In fact, colonic ammonia output is more reduced by a combination of cleansing-sterilization than compared with only one of these methods used separately (19). To investigate the isolated effect of lactose on PSE a study should be done comparing lactose enemas with tap-water enemas. Such a study was planned in our unit, but for ethical reasons our Clinical Investigation Committee did not approve its performance. Lactose enemas are an effective alternative to neomycin, cathartics, or lactulose in the treatment of acute PSE. When deciding whether or not to use lactose enemas, it is not mandatory that patients be lactose intolerant. Lactose enemas are particularly useful in patients with ilei in whom the oral administration of lactose or lactulose is not feasible. Although the population studied is small, lactose enemas were free of side effects. Also, the enemas are inexpensive and easy to prepare (in Mexico, 1 kg of lactose costs $1.50), an important advantage in developing countries such as Mexico.
Vol. 81, No. 1
References 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. 2. Conn 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. Uribe M, Marquez MA, Garcfa Ramos G, et al. Treatment of chronic portal systemic encephalopathy with lactose in lactase defficient patients. Dig Dis Sci 1980;25:92.4-8. 4. Welch JD, Cassidy D, Prigatano GP, et al. Chronic hepatic encephalopathy treated with oral lactose in a patient with lactose malabsorption. N Engl J Med 1974;291:240-1. 5. Hoyumpa AM, Desmond PV, Avant PR, et al. Hepatic encephalopathy. Gastroenterology 1977;76:184-95. 6. Lewis H, Rojas S, Uribe M, et al. Similar effect of lactose and lactulose on in vitro fecal ammonia generation and bacterial flora (abstr). Gastroenterology 1986;78:1206. 7. Vince AJ, Burridge SM. Ammonia production by intestinal bacteria: the effects of lactose, lactulose and glucose. J Med Microbial 1980;13:177-92. 8. Imler M, Kurtz D, Bockel R, et al. Etude comparative du traitment de 1 encephalopathie portocave par le lactulose, les batiles lactiques et les antibiotiques. Therapeutique 1971; 47:237-48. 9. Kersh ES, Rifkin H. Lactulose enemas. Ann Intern Med 1973;78:81-4. 10.Conn HO. Trailmaking and number connection tests in the assessment of mental state in portal systemic encephalopathy. Dig Dis Sci 1977;22:541-50. 11.Uribe M, Farca A, Marquez MA, et al. Treatment of chronic portal systemic encephalopathy with bromocriptine a doubleblind controlled trial. Gastroenterology 1979;76:1347-51. 12. MC Dermont WV, Adams PD, Riddell AG. Ammonia levels in blood and cerebrospinal fluid. Proc Sot Exp Biol 1955;88:3802. 13. Conn HO, Lieberthal MM. Measurement of stool pH. In: Conn HO, Lieberthal MM, eds. The hepatic coma syndromes and lactulose. Chapt 11. Baltimore: Williams & Wilkins, 1979:295305. 14.Rickerms A, Todd HN. Introducci6n a la estadistica. 2nd ed. Barcelona, EspaAa: Editado por la Compaiiia Editorial Continental SA, 1974:90-3. 15. Conn HO, Lieberthal MM. Lactulose enemas. In: Conn HO, Lieberthal MM, eds. The hepatic coma syndromes and lactulose. Chap 13. Baltimore: Williams & Wilkins, 1979:317-22. 16.Bond JH, Jr, Levitt MD. Investigation of small bowel transit time in man utilizing pulmonary hydrogen (H,) measurements J Lab Clin Med 1979;85:546-55. 17.Price JB, Jr, Sawada M, Voorhees AB, Jr. Clinical significance of intraluminal pH in intestinal ammonia transport. Am J Surg 1970;119:595-8. 18. Pertsiounis S, Egger G, Bircher J. Radiological “quantitation” of osmotic effect. Baden bei Wein, Austria: Symposium on Lactulose, 1969. 19. Wolpert E, Phillips SF, Summerskill WHJ. Ammonia production in the human colon. Effects of cleansing, neomycin and acetohydroxamic acid. N Engl J Med 1970;283:159-64. 1.