encephalopathy - NCBI

PATHOPHYSIOLOGICAL CHANG ASSOCIATED WITH INCREASIN GRADE OF HEPATIC

ENCEPHALOPATHY Rami Eliakim, MD, Daniel Shouval, MD, and Marcel Eliakim, MD Jerusalem, Israel

The pathophysiological changes occurring with increasing grade of encephalopathy were examined in 93 consecutive episodes in 44 patients with liver cirrhosis (37 posthepatic). The incidence of gastrointestinal bleeding and leukocytosis increased significantly when the grade advanced from 1 to 5. The following variables showed a trend for change that did not reach statistical significance: rising serum bilirubin, SGOT, and BUN levels; decreasing serum sodium and chloride levels; and increased incidence of infection. The mean values of the following variables were significantly different in 25 fatal episodes and 68 survivors, implicating a bad prognosis: high serum bilirubin, alkaline phosphatase, and BUN levels; low serum albumin, sodium, and chloride levels; and a higher incidence of severe infections (sepsis, infected ascitic fluid). Because increasing grade of encephalopathy is the most important factor in determining the prognosis of hepatic encephalopathy (mortality 0, 10, 5, 19, and 85 percent in grades 1 to 5, respectively), more efforts should be made to understand and prevent the pathophysiological changes associated with advancing grades of encephalopathy.

From the Department of Medicine A, Hadassah University Hospital, Ein Kerem, Jerusalem, Israel. Requests for reprints should be addressed to Dr. M. Eliakim, Department of Medicine A, Hadassah University Hospital, Jerusalem, Israel.

986

The prognosis of patients with hepatic encephalopathy depends primarily on the stage ofthe disease. Thus, in a series of 93 episodes in 44 patients classified in grades 1 to 5,1 the mortality was 0, 10, 5, 19, and 85 percent, respectively.2 In a comparable series of Danish patients,3 which analyzed 100 episodes in 61 patients classified in grades 1 to 3,4 mortality increased from 45 percent in grade 1 to 56 percent and 76 percent in grades 2 and 3, respectively. The following factors were significantly associated with increased mortality: sepsis, deteriorating liver function, gastrointestinal bleeding, uremia, and electrolyte disturbances.2 The present article will attempt to analyze the pathophysiological changes associated with the advancement of the stage of encephalopathy.

PATIENTS AND METHODS Ninety-three episodes of acute hepatic encephalopathy occurring in 44 patients with chronic liver disease hospitalized at the Hadassah University Hospital during the years 1970 to 1979 were analyzed. Patients with primary or metastatic liver tumors, extrahepatic biliary disease, and fulminant hepatitis were not included. Thirty-seven patients had posthepatic cirrhosis, two had Wilson's disease, two had Budd Chiari syndrome, and three had alcoholic cirrhosis, primary biliary cirrhosis, and a,-antitrypsin deficiency, respectively. The diagnosis was based on clinical and laboratory data and confirmed by liver biopsy or autopsy in 33 cases.

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 80, NO. 9, 1988

HEPATIC ENCEPHALOPATHY

The patients were graded in stages 1 to 5' on admission and again during the most advanced stage of the disease. Data extracted from the charts included demographic characteristics, symptoms and signs relative to the central nervous system, liver disease, renal function, precipitating factors, various laboratory examinations, treatment, and outcome. Statistical evaluation was performed by one-way analysis of variance, Student's t, and X2 tests.

PROTHROMBIN

100 7.0

GLg8,2Vl¾ GLOBULIN

4.

X:

g .*

gI.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'

gi.~~.0

ALBUMIN RESULTS Twenty-eight patients were male and 16 were female; the mean age of the group was 49.3 ± 19.6 years. Twelve patients were of Ashkenazi (West and North European) extraction, 24 were of Sephardi (originally Spanish) origin, 7 were Arabs, and 1 was Persian. Nine patients were HBSAg positive. Because analysis of the data of HB,Ag positive and negative patients showed no significant differences, the two groups will be analyzed jointly. Twenty-two patients had had portocaval shunts in the past. The distribution of the episodes by initial grade of encephalopathy was as follows: grade 1, 4 episodes; grade 2, 39; grade 3, 20; grade 4, 21; and grade 5, 9. During the hospitalization, the condition of some of the patients deteriorated and the final distribution by grade was as follows: grade 1, 3 episodes; grade 2, 29; grade 3, 20; grade 4, 21; and grade 5, 20. The grade of encephalopathy was the decisive prognostic factor, the mortality being 0, 10, 5, 19, and 85 percent in grades 1 to 5, respectively (P = .0003). Of the 25 episodes that terminated in death, 17 were in grade 5, 4 in grade 4, 1 in grade 3, and 3 in grade 2. The mean age of the patients (estimated by initial grade of encephalopathy) was 59, 48, 44, 49, and 62 in grades 1 to 5, respectively. The mortality in patients aged 60 years (35 episodes) was 34.3 percent (P > .1). The mortality in 62 episodes involving men was 26 percent, and that in 31 women was 32 percent

(P > .1). The changes in liver function tests by grade of encephalopathy are shown in Figure 1. Serum albumin and prothrombin levels were low, while serum bilirubin, SGOT, alkaline phosphatase, and globulin levels were high in all grades. When examined by one-way analysis of variance, there were no significant

2.0

ALKALINE 200 PHOSPHATASE 100K

L

units

S GOT units

200 0

BILIRUBIN 8.0 mg '1

Lik> 1

2

3

4

5

GRADE Figure 1. Liver function tests by initial grades of hepatic encephalopathy (93 episodes)

differences in the levels of any of these measurements in the different grades of encephalopathy. Nevertheless, the mean levels of bilirubin, alkaline phosphatase, and albumin were significantly different in episodes that terminated in death, compared with episodes of patients who survived (Table 1). Hematological changes in the different grades of encephalopathy are summarized in Figure 2. The hemoglobin level and platelet counts were low but similar in all five grades, while the white cell count rose significantly with increasing grades.


987


TABLE 1. PROGNOSTIC FACTORS IN 93 EPISODES OF HEPATIC ENCEPHALOPATHY Termination of Episodes

Survival (n = 68)

Factor

Death (n = 25)

21 19 Gastrointestinal bleeding 11 Electrolyte disturbance* 15 11 Infection** 9 7500 ± 5879 (64)*** White cell count/mm3 7228 ± 1400 (24) 4.7 ± 4.8 (68) Serum bilirubin, mg% 14.5 ± 13.2 (24) Serum albumin, g% 2.48 ± 0.56 (64) 2.03 ± 0.66 (21) 172 ± 99 (62) Alkaline phosphatase, ,u/mL 240 ± 165 (20) 27 ± 13.5 (64) 62 ± 32 (25) BUN, mg% 105 ± 8.7 (59) 92 ± 11.5 (19) Serum chloride, mEq/L 134 ± 7.5 (66) 127 ± 10.4 (24) Serum sodium, mEq/L . * Serum sodium 130, K < 3.5 mEq/L ** Sepsis or infected ascitic fluid *** Mean ± SD (number of episodes in which the test was performed is indicated in parentheses)

N.S.

3000

PLATELETS3000 mm3

200,000[ 100.000

w B c

15000r ,12,000 Pp . 1). However, severe infections (sepsis and infected ascitic fluid) were significantly more frequent in fatal episodes (Table 1). Electrolyte disturbances were also more frequent in fatal episodes than in survivors (60 and 16 percent, respectively, P < .003). The mean value of variables that were not statistically different in affecting the outcome ofeither survival or death are listed in Table 2. Of interest is the level of SGOT, which was higher in fatal episodes but not to the extent of statistical significance, and that of ammonia, which was similar in both groups.



B UBUNN mg I*

ASCITES 6/0

~~5030L

1oL>

BICARBONATE 30

a,

CHLORIDE l10j mEql l 100

100

N.S. 50k-NS

ELECTROLYTE 100 N.S. DISTURBANCE 50 01.0

90L

POTASSIUM 5.0

mEq/I

INFECT ION .1./

100 - N.S. 50[-

61 BLEEOING

1X0 K1

'1/

3.0 L

100

1 p=0.03

10

50n l

mSODqIU 11.0mEqIl

130 LF7

2

3 4 GRADE

S

Figure 4. Precipitating factors in different grades of

I

encephalopathy (93 episodes)

GRADE Figure 3. Renal function and electrolytes by initial grades of encephalopathy (93 episodes)

DISCUSSION The present study confirms the well-established fact that the prognosis of hepatic coma is related primarily to the degree of neuropsychiatric disturbances, that is, to the grade of encephalopathy. The importance of the pathophysiological changes associated with increasing stages of encephalopathy lies therefore in the possibility of arresting the advancement of the disease to higher stages. Analysis of the study data shows that the incidence of gastrointestinal bleeding and increasing leukocytosis, presumably indicating infection, were significantly associated with an advancing grade of encephalopathy. In addition, the following variables showed a trend for change that, however, did not reach significant levels when examined by one-way analysis of variance: increasing serum bilirubin, SGOT, and BUN, and decreasing serum sodium and chloride levels. There was also a trend to-

ward increased incidence of infection and electrolyte disturbances. Because the small number of cases in grades 1 and 5, as well as the wide standard deviation, may have prevented the attainment of statistical significance by some of these measurements, an attempt was made to confirm their prognostic role by comparing their mean values in patients who died with those who survived. Twenty-one of the 25 fatal episodes occurred in encephalopathy of (final) grades 4 and 5, and the total mortality in these two grades was 51 percent (21 out of 41). On the other hand, the mortality in grades 1 to 3 was 7.7 percent (4 out of 52). The values of the following variables were found to be significantly different in fatal and surviving cases, thereby implicating a bad prognosis: higher incidence of electrolyte disturbances and severe infection; high serum bilirubin, alkaline phosphatase, and BUN levels; and low serum albumin, sodium, and chloride levels (Table 1). Thus, the significance of most of the trends found in the analysis by grade of encephalopathy was confirmed by comparing the mean values in survivors and fatal cases. Increasing grades of encephalopathy are therefore


989


TABLE 2. VARIABLES WITH NO PROGNOSTIC SIGNIFICANCE FOR THE OUTCOME OF HEPATIC ENCEPHALOPATHY (P> .1)

Termination of Episodes

Survival (n = 68)

Factor

Hemoglobin, g% Platelets/mm3

SGOT, As/mL

Prothrombin time, % of normal Serum globulin, g% Arterial NH3, 14g% Serum B12, pg/mL Serum potassium, mEq/L Serum HCO3, mEq/L Serum calcium, mg%

10.5 ± 1.9 (67)* 89,000 ± 49,000 (57) 190 ± 144 (64) 37 ± 18 (65) 4.12 ± 0.68 (64) 162 ± 132 (34) 1,939 ± 888 (33) 4.15 ± 0.86 (66) 21.8 ± 5.6 (32) 8.3 ± 0.97 (42)

Death (n = 25)

10.4 ± 93,000 ± 274 ± 29 ± 4.28 ± 176 ±

2,264 ± 4.46 ± 19.8 ± 8.0 ±

2.3 (24) 79,000 (22) 296 (21) 12 (21) 0.94 (21) 100 (13) 899 (12) 1.37 (24) 8.2 (13) 0.55 (9)

* Number of episodes in which examination was performed is indicated in parentheses

associated with a rising incidence of gastrointestinal bleeding, severe infections, and deteriorating liver and renal function. It is of interest that ascites was not identified as a risk factor for the outcome of a single episode of encephalopathy. This is in contrast with previous reports3'5 and may have been the result of more vigorous diuretic therapy in the study patients, as reflected in the frequent occurrence of electrolyte changes. This, of course, does not preclude the significance of ascites in the long-term prognosis of the cirrhotic patient. Also of interest is the finding that "minor" infections played no prognostic role in the outcome of encephalopathy, and that only "severe" infections (sepsis, infected ascites) were of prognostic importance. Prytz and Sloth3 reported that older patients had a significantly worse prognosis, a finding that could not be confirmed in the present study. The finding of similar arterial ammonia levels in episodes ending in either survival or death was surprising in view of the expected rise with increasing encephalopathy.6 This finding is, however, in keeping with previous reports of a lack of correlation between blood ammonia and the depth of coma.7 Factors that have received less attention are the etiology of cirrhosis and the previous performance of shunt operations. While most European and American series are composed of alcoholic cases, the majority of the study patients had posthepatitic cirrhosis. The authors have previously reported2 that the threeyear survival after the first episode of encephalopathy 990

in a series was superior (30 percent) to that in other series,3'5 possibly because of the different etiology of the disease.2 It is also possible that alcoholic etiology might influence differently the course of an episode of encephalopathy. To the best of the authors' knowledge, there are no controlled studies on the effect of different etiologies on the pathophysiological changes in hepatic encephalopathy, or studies comparing the course of encephalopathy in patients with or without portocaval shunts. The therapeutic implications of the present findings are obvious. Although they do not offer new ways of treatment, the importance of preventing changes associated with advancing grades of encephalopathy is clearly demonstrated. Thus, restraint from aggressive therapy with steroids, diuretics, catheters, and so on may be of great value in the prevention of bleeding, infection, and electrolyte disturbances. Immediate correction of anemia and electrolyte and acid-based changes, and treatment of infection are of major importance. Suppression of production of potential false transmitters by neomycin and/or lactulose are routine and will not be discussed. An attempt to arouse the confused or comatose patient with L-dopa8'9 should be made in every case (unless clearly contraindicated), so as to decrease the necessity of intravenous infusions, catheters, and stomach tubes as possible sources of infection or bleeding. Most of all, increased efforts to understand and correct the bleeding tendency are of crucial importance.



Literature Cited

1. Sherlock S. Diseases of the Liver and Biliary System, ed 5. Oxford and Edinburgh: Blackwell Scientific Publications, 1975. 2. Eliakim R, Eliakim M. Prognostic factors in hepatic encephalopathy: A review of 93 episodes. Harefuah 1982; 102: 409-413. 3. Prytz H, Sloth K. Hepatic coma in cirrhosis of liver-The course and prognosis in 100 consecutive cases. Scand J Gastroenterology 1973; 8:229-233. 4. Sherlock S. Diseases of the Liver and Biliary System, ed 4. Oxford and Edinburgh: Blackwell Scientific Publications, 1968.

5. Sherlock S, Summerskill WHJ, Dawson AM. The treatment and prognosis of hepatic coma. Lancet 1956; 2:689-694. 6. Conn HO, Lieberthal MM. The Hepatic Coma Syndromes and Lactulose. Baltimore: Williams and Wilkins, 1979. 7. Phear EA, Sherlock S, Summerskill WHJ. Blood ammonium levels in liver disease and hepatic coma. Lancet 1955; 1:836-

849. 8. Lunzer M, James IM, Weinman J, et al. Treatment of chronic hepatic encephalopathy. Gut 1974; 15:5555-5561. 9. Chajek T, Berry EM, Friedman G, et al. Treatment of acute hepatic encephalopathy with L-dopa. Postgrad Med J 1977; 53: 262-265.

-I~~~~~~~~~~~~~~~~~~E a

S.~~~~~~~~~~~~~~~~~~~~~~~~~~..

*.

..

r~~~~~~~~~~~~~~~~~~~~~~~~J

FiberOon is the firs fiber'laxative in

convenient easy-to-swallow tablets. Unlike Metamucil® and other

* 11 | Q . !

measuring, and no complaints about *~~~~,. ~~~taste with FiberCon. And while Metamucil bulks in the stomach, there's less chance of stomach bloating with FiberCon because its bulking action is primarily in the intestine.

- -:

Metamucil isa registered trademarkoftProcter &

Gamble.

powders, there's no mixing, no

\

u\ci- - Available in packages of 36 and 60 tablets

.

1X.

dtA Calcium polycarbophil

Bulk-forming fiber laxative

Lederle Laboratories, A Division of American Cyanamid Company, Wayne, New Jersey 07470 The fiber action symbol is a registered c 1988 Lederte Laboratories Printed in USA January1988 Y 094-8C trademark of Lederle Laboratories.